-
Active Hedgehog Signaling Regulates Renal Capsule
Morphogenesis
by
Hovhannes Martirosyan
A thesis submitted in conformity with the requirements for the
degree of Master of Science
Laboratory Medicine and Pathobiology University of Toronto
© Copyright by Hovhannes Martirosyan 2013
-
ii
Active Hedgehog Signaling Regulates Renal Capsule
Morphogenesis
Hovhannes Martirosyan
Master of Science
Laboratory Medicine and Pathobiology University of Toronto
2013
The renal capsule is a flattened layer of cells which surround
the kidney. Expression of
the transcription factor Foxd1 is required for normal
development of the capsule.
Furthermore, current evidence suggests that during development
the capsule
progenitors are in a state of active hedgehog signaling. We
hypothesize that hedgehog
plays a role in modulating capsule morphogenesis in the
embryonic kidney. To test the
hypothesis hedgehog signaling was inhibited in the capsule via
Foxd1Cre mediated
deletion of Smoothened (Smo), the activator of the pathway.
Mutant kidneys were
approximately 48% smaller in volume and had a 42% decrease in
nephron number.
Furthermore, mutants displayed abnormal patterning of the
capsule where regions on
the surface of the kidney had no capsule cells. The
discontinuous capsule phenotype
was observed only after E13.5. Additionally, capsule cells
progressively lost expression
of known markers Foxd1 and Raldh2 and their proliferative
capacity was decreased by
54% at E13.5.
-
iii
Acknowledgments
I would like to express a great deal of gratitude to Dr. Norman
Rosenblum for the
supervision he has provided me over the last 2 years. Not only
did he give me the
opportunity to be a part of his wonderful lab and have a chance
to work on such a new
and exciting project he was also the best scientific mentor I
have ever had. Thanks to
Norm’s guidance and mentoring I had a successful grad school
experience and more
importantly I am now a better scientist able to communicate my
research and
knowledge with ease and clarity.
An extension of my gratitude goes out to my supervisory
committee, Dr. Michael Ohh
and Dr. Herman Yeger for their time and mentorship. They have
encouraged me to
continually ask questions and search for the bigger picture in
all scientific research.
The group of individuals I was surrounded with everyday in the
lab was the best
collection of scientists I could have asked for. Their constant
assistance in the lab
setting as well as emotional help during times of trouble was
irreplaceable. Not to
mention the wealth of knowledge they all provided me whenever I
needed it, I could not
have finished my degree without it. I would like to express
gratitude to Dr. Jason Cain,
Dr. Lin Chen, Dr. Lijun Chi, Dr. Valeria Di Giovanni, Josh
Blake, Meghan Feeny,
Tayyaba Jiwani, Jinny Kim, and Joanna Smeeton. I also want to
thank Doug Holmyard
at the Advanced Bioimaging Centre for SEM imaging assistance.
Finally, I would like to
acknowledge the SamuelLunenfeld Research Institute's CMHD Mouse
Physiology
Facility for their technical screening services
(www.cmhd.ca).
Of course, last but not least I would like to thank my
girlfriend Carly Willemsma for her
love and support. She gave me many many words of encouragement
that gave me the
strength to finish my Master of Science degree.
I am but a seed Exploring the Milky Way, It taught me to
flower.
-
iv
Table of Contents
Acknowledgments
..........................................................................................................................
iii
Table of Contents
...........................................................................................................................
iv
List of Tables
................................................................................................................................
vii
List of Figures
..............................................................................................................................
viii
List of Abbreviations
......................................................................................................................
x
1. Chapter One: Introduction
..........................................................................................................
1
1.1 Overview
.............................................................................................................................
1
1.2 Mouse Kidney Development
..............................................................................................
2
1.2.1 Specification of metanephric mesenchymal
........................................................... 2
1.2.2 Ureteric bud outgrowth and branching
...................................................................
2
1.2.3 Nephrogenesis
.........................................................................................................
5
1.2.4 Development of the ureter
......................................................................................
7
1.3 Stroma
.................................................................................................................................
8
1.3.1 Introduction
.............................................................................................................
8
1.3.2 Maintenance of outer and inner differentiation zones
.......................................... 10
1.3.3 Stromal cells in branching morphogenesis
........................................................... 12
1.3.4 Vitamin A in branching morphogenesis and stromal cell
patterning ................... 12
1.4 Hedgehog Signaling
..........................................................................................................
14
1.4.1 Sonic, Indian, and Desert hedgehog
.....................................................................
14
1.4.2 Signaling mechanism
............................................................................................
14
1.5 Hedgehog Signaling is Required for Kidney Development
............................................. 17
1.6 Stromal Cells are in State of Active Hedgehog Signaling
................................................ 19
1.7 Rationale
...........................................................................................................................
21
1.8 Hypthesis
...........................................................................................................................
21
-
v
1.9 Model
................................................................................................................................
21
2 . Chapter Two: Methods
.............................................................................................................
23
2.1 Mouse Breeding
................................................................................................................
23
2.2 Tissue Preparation
.............................................................................................................
24
2.3 β-galactosidase
Staining....................................................................................................
24
2.4 Histology and Immunofluorescence
.................................................................................
25
2.5 Capsule Cell Proliferation Analysis
..................................................................................
25
2.6 In situ mRNA Hybridization
.............................................................................................
26
2.7 Kidney Volume Analysis and Total Nephron Number Count
.......................................... 26
2.8 Scanning Electron Microscopy
.........................................................................................
27
2.9 Blood and Urine Collection
..............................................................................................
27
2.10 Data Analysis
....................................................................................................................
28
3 . Chapter Three: Results
.............................................................................................................
28
3.1 Smo null mutant kidneys are characterized by decreased
expression of HH target gene mRNA
...............................................................................................................................
28
3.2 Patterning of the outer renal cortex is abnormal in Smo null
mutant kidney .................... 29
3.3 Smo null mutant capsule appears discontinuous after E13.5
............................................ 33
3.4 Smo null mutant kidneys are smaller in volume and have fewer
nephrons than
wildtype kidneys
...............................................................................................................
37
3.5 Smo null mutant kidneys have decreased Foxd1 and Raldh2
expression ......................... 37
3.6 Proliferative capacity of mutant capsule cells is
significantly decreased ......................... 42
3.7 Mutant mice do not always have a discontinuous capsule
............................................... 44
3.8 Viable mutants display normal kidney function
...............................................................
46
3.9 Gli3 deficiency improves renal patterning and capsule
formation ................................... 49
4 . Chapter Four: Discussion and Future Experiments
..................................................................
49
4.1 Deleterious effects on capsule formation by decreased
hedgehog signaling .................... 52
4.2 Capsule formation is required for viability
.......................................................................
53
-
vi
4.3 Effects of loss of capsule on renal patterning
...................................................................
54
4.4 Gli3 repressor influences renal capsule development
....................................................... 56
4.5 Future Directions
..............................................................................................................
56
5 . Chapter Five: Conclusion
.........................................................................................................
58
6 . Chapter Six: References
...........................................................................................................
60
-
vii
List of Tables
Table 1: Urinalysis of adult wildtype and Smo null littermate
mice. ................................... 48
-
viii
List of Figures
Figure 1. Overview of kidney development initiation
..............................................................
3
Figure 2. Development of the metanephric kidney
.................................................................
4
Figure 3. Stromal cells express Foxd1 and Raldh2.
...............................................................
9
Figure 4. Foxd1 null mutant kidneys
.......................................................................................
11
Figure 5. Vitamin A in branching morphogenesis and stromal cell
patterning ................. 13
Figure 6. Overview of hedgehog signaling
.............................................................................
16
Figure 7. Hedgehog signaling in kidney development
......................................................... 18
Figure 8. Genetic elimination of Ptc1 resulting in hydropelvis
and death immediately
following birth
..............................................................................................................................
20
Figure 9. Stromal cells are characterized by active hedgehog
signaling .......................... 22
Figure 10. In situ hybridization of Gli1 at
E13.5.....................................................................
30
Figure 11. In situ hybridization of Ptc1 at E13.5
....................................................................
31
Figure 12. Hematoxylin and eosin (H&E) stained P0 kidneys
............................................ 32
Figure 13. Scanning Electron Microscopy (SEM) images of E16.5
wildtype and Smo null
kidneys
.........................................................................................................................................
34
Figure 14. H&E sections of wildtype and Smo null kidneys at
various embryonic stages
.......................................................................................................................................................
35
Figure 15. SEM images of E13.5 wildtype and Smo null kidneys
...................................... 36
Figure 16. Kidney volume and nephron number analysis of E18.5
kidneys ..................... 38
Figure 17. In situ and immunofluorescence of Foxd1 at various
stages ........................... 40
../../../Dropbox/John/Thesis%20edits%20for%20submission_NRJan3.doc#_Toc345342789
-
ix
Figure 18. In situ hybridization of Raldh2
...............................................................................
41
Figure 19. A ratio of BrDU positive (proliferative) capsule
cells over total number of
capsule cells was used as an index of proliferative capacity
.............................................. 43
Figure 20. H&E sections of wildtype, viable mutant, and
non-viable mutant P0 kidneys 45
Figure 21. Blood biochemistry of adult wildtype and Smo null
littermate mice ................ 47
Figure 22. H&E sections of wildtype, Smo null mutant, and
Smo null; Gli3 heterozygous
deleted mutant P0 kidneys
.......................................................................................................
50
Figure 23. H&E sections of control, Smo null mutant, and Smo
null; Gli3 null mutant
E18.5 kidneys
.............................................................................................................................
51
-
x
List of Abbreviations
BrDU Bromodeoxyuridine
E Embryonic day
Gdnf Glial-derived growth factor
GCPS Greig cephalopolysyndactyly syndrome
Hh Hedgehog
ISH in situ hybridization
MET Mesenchymal-to-epithelial
PHS Pallister-Hall syndrome
PFA Paraformaldehyde
Ptc Patched
PAS Periodic acid-Schiff
RA Retinoic acid
Raldh2 Retinoic acid dehydrogenase-2
Rars Retinoic acid receptors
SEM Scanning electron microscopy
Smo Smoothened
SA Surface area
UB Ureteric bud
-
1
1. Chapter One: Introduction
1.1 Overview
Kidneys are bilateral organs located in the posterior abdominal
cavity. Their function is
essential for maintaining body fluid composition and
homeostasis, filtering blood, and
excreting waste products. The nephron is the basic functional
unit of the kidney. It is
composed of a glomerulus, where plasma is filtered from
capillaries into Bowman’s
capsule, the proximal tubule, which reabsorbs nutrients and
electrolytes from the filtered
load, the loop of Henle, where surrounding tissue and filtrate
increase osmolality to
concentrate urine, and the distal convoluted tubule, which
reabsorbs water and sodium.
The nephron is connected to the collecting duct, which collects
urine and excretes it into
the bladder. Kidney development is initiated during the 5th week
of gestation in humans
(at embryonic day 10.5 in mice). Unlike other organs, formation
of the kidney includes
three phases during which transient and rudimentary kidney
structures form and
degenerate until the metanephros gives rise to the complex and
functional metanephric
kidney. A fully formed kidney is composed of an expansively
branched collecting duct
system and a large number of nephrons ranging ~13000 (±1300) in
mice and ~650000
(±200000) in humans2-4. The proper development of this complex
organ relies on cell
fate decisions, cell migration, cell differentiation, and
organization of cells into three-
dimensional structures. Cell communication is known to be
central in nephrogenesis
and ureteric bud branching. Four families of signaling molecules
have been implicated
in renal cell communication – fibroblast growth factors, wnt
proteins, TFG-beta super
family factors, and hedgehog proteins. A thorough understanding
of molecular signaling
-
2
underlying renal development will be crucial for elucidating the
pathogenesis of human
kidney disorders and developing therapeutic approaches.
1.2 Mouse Kidney Development
1.2.1 Specification of metanephric mesenchyme
The mammalian kidney is derived from the intermediate mesoderm,
tissue from a
neurula-stage embryo that forms the urogenital and reproductive
system5. Intermediate
mesoderm gives rise to nephric precursors which detach,
differentiate, and reorganize
to form an epithelial tube via a mesenchymal-to-epithelial (MET)
transition6,7. This tube,
also termed the nephric duct, grows caudally until it is
adjacent to the metanephric
mesenchyme (Figure 1). The specification of the metanephric
mesenchyme is
independent of the presence or absence of the nephric duct.
Transcription factors Eya1,
Six1/Six4, and Odd1 are the first regulators responsible for
specifying the metanephric
mesenchyme from the caudal end of the intermediate mesoderm8.
Odd1 is an early
molecular marker that acts in the metanephric mesenchyme
upstream of Eya1 and
Six1/4 to promote its formation and survival9. Eya1 and
Six1/Six4 activity is important for
upregulation of downstream genes. Eya1 and Six1/Six4 deficient
embryos lose Gdnf
expression and the ureteric bud (UB) fails to invade the
mesenchyme10,11.
1.2.2 Ureteric bud outgrowth and branching
At E10.5 of mouse embryonic development Glial-derived growth
factor (Gdnf) is
secreted by metanephric mesenchyme cells and binds to Ret
tyrosine kinase receptor
expressed on the surface of nephric duct cells12-14. In
response, a subset of duct cells
emerge to form the UB, invade the mesenchyme and begin branching
(Figure 2A)15.
Reciprocal signaling interactions between the UB and the
metanephric mesenchyme
-
3
Figure 1. Overview of kidney development initiation. At
embryonic day 8.5 mouse
kidney development begins with the formation of the nephric
duct. The nephric duct
extends caudally and fuses with the urogenital sinus. At E10.5,
the ureteric bud forms
adjacent to the metanephric mesenchyme. Reciprocal signaling
stimulates the ureteric
bud to invade the mesenchyme and branch. At E11.5 a T-shaped
structure is observed.
Mesonephric tubules rostral to the UB are not shown (adapted
from Davidson et al.
2008)1.
-
4
Figure 2. Development of the metanephric kidney. (A) Ureteric
bud outgrowth is
dependent on Gdnf and its receptor Ret. Gdnf released from the
metanephric
mesenchyme binds to Ret receptors on the nephric duct
stimulating an outgrowth which
invades and continues to branch within the mesenchyme. (B)
Nephrogenesis initiates
when cells surrounding the ureteric bud tips begin to condense
and form pretubular
aggregetes. The aggregates convert into renal vesicles and
further mature into comma-
shaped and S-shaped bodies. The S shaped body fuses to the
collecting duct and
forms mature nephron when a glomerulus develops and the proximal
tubules elongate
and grow towards the medulla(adapted from Gilbert, 2006 and
Davidson et al. 2008)1,5
-
5
result in two key processes: ureteric branching morphogenesis
and formation of
metanephric derived nephrogenic precursors. Following invasion
of the metanephric
mesenchyme by the UB, expression of Ret becomes restricted to UB
tips while Gdnf
expression gets limited to the condensing mesenchyme surrounding
the tips14,16. Gdnf-
Ret signaling triggers further generations of branching
morphogenesis with eventual
formation of the collecting system of the kidney15 Gdnf
expression must be tightly
regulated in the metanephric mesenchyme with both transcription
activators (Eya1,
Sall1, Pax2) and repressors (Slit2, Robo2) to allow normal UB
invasion and branching
without induction of ectopic UBs from portions of the nephric
duct anterior to the normal
site17-20. Furthermore, the UB induces nephrogenesis, thus there
is a strong correlation
between the number of UB branches and the number of nephrons in
a kidney.
Disturbance of any major component of reciprocal signaling
between the UB and the
metanephric mesenchyme that affects renal branching
morphogenesis can cause
decreased nephron number in addition to aplasia (no kidney) or
dysplasia (abnormal
kidney tissue)21,22. In the absence of renal replacement therapy
immediately, renal
aplasia is lethal postnatally and dysplasia is the leading cause
of chronic renal failure in
children23.
1.2.3 Nephrogenesis
Invasion of the metanephric mesenchyme by the UB at E11 induces
mesenchymal cells
to condense or rearrange in a cap-like formation around the UB
tip 4 – 5 cell layers thick
that are morphologically distinguishable from more peripheral
‘uninduced’ mesenchyme.
The newly formed cap-mesenchyme is morphologically distinct from
uninduced
mesenchymal cells and expresses genes such as Eya1, Sall1, Pax2,
Six2, and secreted
factors Gdnf and Bmp7. The cap is the source of progenitors
which differentiate into
-
6
nephrons. Constant proliferation of these cells is critical to
proper kidney development
and nephrogenesis. Signals within the cap mesenchyme promote
further condensation
around UB tips, promote survival by inhibiting apoptosis, and
promote proliferation to
maintain a population of progenitor cells that will become
nephrons24-26. Cap-like
formation of mesenchymal cells appears dependent on Smad4
encoding an intracellular
mediator of TGF-β/BMP signaling, as conditional knockout of
Smad4 in the metanephric
mesenchyme results in mesenchymal cells failing to coalesce
around UB tips26,27.
Increased apoptosis of mesenchymal cells is observed in mice
deficient in Pax2
demonstrating an important role for Pax2 in metanephric
mesenchyme survival28. Bmp7
function is also required for metanephric mesenchyme survival as
the cells that form a
cap around the UB undergo apoptosis at E13.5 in Bmp7 mutants29.
Proliferation of the
cap mesenchyme is promoted by Six2, a fate mapping study has
shown Six2-
expressing cap mesenchyme cells can proliferate and expand their
numbers to
accommodate for numerous rounds of branching and
nephrogenesis25.
Following cap mesenchyme surrounding the UB, nephron formation
begins at E11.5.
Clusters of mesenchymal cells appear on either side of the UB
tip and proliferate to
become aggregates of 30 cells or more30. As the UB branches and
grows the pre-
tubular aggregates become located beneath the UB tips. At E12.5
the aggregates
undergo MET to form renal vesicles which undergo further
proliferation and express a
variety of genes to give rise to comma shaped bodies and then
S-shaped bodies
(Figure 2B). The molecules which control nephrogenesis include
Wnt9b and Wnt4
which are responsible for pretubular aggregate and renal vesicle
formation,
respectively27,31. Wnt9b is expressed throughout the collecting
system but it only
activates aggregate formation in cap mesenchyme adjacent to the
UB tips due to
-
7
regulatory factors such as Six2 which inhibits pretubular
aggregate formation by Wnt9b
in the remainder of the cap mesenchyme25,32. Unlike Wnt9b which
is a paracrine
molecule, Wnt4 is expressed in aggregates and acts in an
autocrine manner to
propagate transition to renal vesicles33. Wnt9b also activates
expression of Wnt4, Fgf8,
and Pax8 in the pretubular aggregate, and then Wnt4 maintains
the expression of these
genes and induces Lim134-36. Conditional inactivation of Lim1 in
the cap mesenchyme
has demonstrated a requirement of Lim1 for the progression of
the renal vesicle to the
comma-shaped body stage35,37. Mechanisms controlling further
segmentation of
nephrons are not completely clear. Some data indicates that
Notch signaling is involved
in proximal-distal patterning to promote the formation of
podocytes, proximal tubules,
and the loop of Henle38-40.
1.2.4 Development of the ureter
The ureter which is derived from the initial portion of the UB
is the tube which propels
urine from kidneys to the bladder. It is a multi-layered
structure and it uses peristaltic
machinery to function. Inductive signaling from tail bud
mesenchyme to the UB initiates
maturation of the first UB segment to form the ureter. At E12.5
the distal end of the UB
separates from the nephric duct41. Inductive signaling from ret
expression, retinoic acid
receptors α/β, and vitamin A mediate formation of a connection
between the ureter and
the bladder42. At E15.5 the epithelium of the ureter
differentiates into the urothelium
giving the ureter the capacity to resist the toxicity of the
urine produced from E16.5
onwards43. The mesenchyme surrounding the ureter originates from
tail bud
mesenchyme and it is marked by transcription factor T-box 18
(Tbx18) as early as
E11.544. This mesenchyme differentiates into smooth-muscle cells
that form layers with
longitudinal and transverse orientation at E14.545. Deficiency
in Tbx18 leads to a
-
8
decrease in Bmp4 expression and results in defective
differentiation of smooth-muscle
cells44. The peristaltic machinery required to move urine from
the pelvis to the bladder
gets established at the ureter-pelvis junction by E18.5 ensuring
full functionality at
birth44.
1.3 Stroma
1.3.1 Introduction
Renal stroma is a small population of cells in the metanephric
mesenchyme that in
contrast to other metanephric cells are not nephrogenic
progenitors. The stromal
population can be broken down into 3 layers which are clearly
visible by E14.5: the
renal capsule (a continuous layer of 1 to 3 cells thick,
flattened and localized to the edge
of the kidney), cortical stroma (cells surrounding the cap
mesenchyme and UB tips),
and medullary stroma46. The capsule and cortical stroma
selectively express Foxd1 and
Raldh2 (Figure 3)47. Originally, the capsule was thought to have
mainly a supportive role
in renal development48. More recently, data has surfaced
implicating FoxD1 and
capsule formation in spatial and temporal patterning of
nephrons47.
It has been reported that Foxd1-expressing cells are initially
observed in a concentrated
cap-like cluster that is localized just anterior to the
metanephric mesenchyme. As
development progresses, these cells are observed in
progressively posterior positions,
becoming integrated into the periphery of the kidney to form the
cortical stroma49.
Deletion of Foxd1 in mice results in renal malformations
including impaired branching
morphogenesis and nephron differentiation and defects in renal
capsule morphology50.
Furthermore, the defects in capsule formation prevent kidneys
from detaching from the
-
9
Figure 3. Stromal cells express Foxd1 and Raldh2. (A) E14
embryonic kidney. The ureteric bud and bud tips are green, cortical
and medullary stroma is pink (B) High magnification of the boxed
area in (A). The capsule is visible as a flattened layer of cells
surrounding the kidney. (C) and (D) Foxd1 and Raldh2 expression in
cortical stroma (stc). Foxd1 is restricted more to capsule cells
(adapted from Levinson et al. 2003)51.
-
10
body wall resulting in fused kidneys which remain in the pelvis
rather than ascending to
the lumbar region (Figure 4AB)47,50. Unlike wildtype kidneys,
without Foxd1 the capsule
layer surrounding the kidney is thicker and populated with
ectopic Bmp4-expressing
cells (Figure 4CD)47. Improper signaling from these cells causes
aberrant patterning in
the nephrogenic compartment which results in mispatterning of
the ureteric tree and
delayed and disorganized nephrogenesis47.
1.3.2 Maintenance of outer and inner differentiation zones
The nephrogenic zone is the domain beneath the renal capsule and
is defined as the
region in which branching of the ureter and nephron formation
occurs. The nephrogenic
zone is located between a more interior layer of differentiating
nephrons, branches and
medullary stroma and an outer layer of renal capsule cells. In
mice the nephrogenic
zone is active until the first week of post-natal life. It is
unknown how the boundaries of
the nephrogenic zone are established and maintained. The
presence of two different
environments on either side of the zone suggests that extrinsic
factors released from
the capsule on cortical side and the more differentiated cell
types on the medullary side
might determine the boundaries. Similar examples can be found in
the gut and genital
tract where stromal mesenchyme generates signals that control
differentiation of their
underlying epithelia into diverse cell types for various
functions52-54. This data supports
the possibility that capsule signaling is important for kidney
development of cell types
adjacent to the capsule or even deep within the organ.
-
11
Figure 4. Foxd1 null mutant kidneys. (A) E16.5 wild-type kidney
displaying glomeruli
(gl) and proximal tubules (pt) in the juxtamedullary region,
just below the nephrogenic
zone. (B) Mutant E16.5 Foxd1 null kidneys are fused and lack a
discrete nephrogenic
zone. (C) and (D) Bmp4lacZ expression is present in the midline
of both Foxd1 null and
heterozygous tissue at E12.5 but there is a wider distribution
of expression around the
kidneys in the mutant. Expression of GFP and lacZ did not
overlap. (E) At E18.5
wildtype branches of the ureteric bud (ub) have bifurcated
numerous times. (F) In the
mutant the branches elongate but bifurcate very infrequently
(adapted from Levinson et
al. 2005)47.
-
12
1.3.3 Stromal cells in branching morphogenesis
FoxD1 is required for correct morphology and cellular
composition of the renal
capsule47. Foxd1 null mutants have a disorganized nephrogenic
compartment as a
result of an improperly matured capsule layer47. Furthermore,
FoxD1 null mutant mice
also display abnormal ureteric bud branching47. The mutant
kidneys display a decrease
in the number of branches as well as a change in branch
patterning. The mutant
branches become severely elongated with few ampula near the
surface suggesting that
mutant branches lost the ability to bifurcate, but not elongate
(Figure 4EF)47. Kidney
rudiments that are treated with Bmp4 display a similar branching
phenotype supporting
the concept that abnormal patterning of the nephrogenic zone in
Foxd1 null mutants
was caused by ectopic Bmp4-expressing cells present in the renal
capsule55.
1.3.4 Vitamin A in branching morphogenesis and stromal cell
patterning
Vitamin A is required for morphogenesis of kidneys as well as
most other fetal
tissue56,57. Vitamin A, also known as retinol, is ingested via
the diet and is irreversibly
oxidized into retinoic acid (RA). RA is a secreted molecule and
its signaling is mediated
by retinoic acid receptors (Rars). The RA-synthesizing enzyme
found in the kidney is
retinoic acid dehydrogenase-2 (Raldh2), the expression of which
is restricted to the
renal capsule and cortical stroma58. Recent studies have
demonstrated a key role for
Vitamin A in renal branching morphogenesis16,59,60. Kidney
explants cultured in the
absence of RA display a significant decrease in branching61.
Kidneys with impaired
branching pattern also lose expression of Ret in the UB tips
(Figure 5A-D)61. Consistent
with these observations, mice deficient in retinoic acid
receptors a and b2 (Rarab2)
display loss of Ret expression and impaired ureteric
branching16,59. Interestingly,
-
13
Figure 5. Vitamin A in branching morphogenesis and stromal cell
patterning.
Hoxb7-GFP kidney rudiments cultured in serum-free medium with
(A) and without (B)
added retinoids. Ret is strongly expressed in the ureteric bud
tips of rudiments cultured
with retinoids, and downregulated in those grown without added
retinoids. (C) and (D)
GFP expression in the ureteric bud of E12 Hoxb7-GFP kidneys.
Embryonic kidneys
grown with retinoids had significantly more ureteric bud
branches than counterparts
grown in the absence of retinoids. (E), (F), and (G) Hoxb7-Ret
transgene rescues renal
development in Rarab2 mutants. (A) Wildtype kidneys have normal
nephrogenic zone
and developing glomeruli. (B) Rarab2 mutant kidneys have few
branches or glomeruli.
(C) Rarab2 mutant with Ret expressed in the ureteric bud
(Hoxb7-Ret transgene) is
rescued and looks morphologically similar to wildtype (adapted
from Levinson et al.
2003)51.
-
14
Rarab2 deficient kidneys are also characterized by a thick outer
capsule. In Rarab2
mutant kidneys where Ret expression was forced in the ureteric
bud, all malformations
were rescued including the stroma surrounding the kidney (Figure
5E-F)62. This
suggests that Ret expression is important for not only branching
morphogenesis of the
kidney but for normal stromal patterning as well. In summary,
Vitamin A is converted
into RA in the renal capsule and gets secreted to the rest of
the kidney where it
upregulates Ret expression in UB tips. Ret signaling from the UB
then induces
branching morphogenesis and supports cortical stromal cell
patterning.
1.4 Hedgehog Signaling
1.4.1 Sonic, Indian, and Desert hedgehog
The hedgehog signaling pathway is highly conserved and controls
tissue
morphogenesis during embryogenesis63. First discovered in 1980
in Drosophila
malanogaster, the hedgehog gene (hh) was found to encode a
secreted protein with an
important role in segment polarity64. There are three proteins
in the mammalian
hedgehog signaling pathway, Sonic (Shh), Indian (Ihh), and
Desert (Dhh). Shh is most
studied for its role in patterning of the neural tube. Ihh is
very active in skeletal bone
formation while Dhh activity seems restricted to the
testis65-68. The hedgehog ligand
functions as a morphogen, a molecule which establishes a
concentration gradient and
affects cell differentiation differently depending on the
location of a cell in reference to
the gradient. Of the three members Shh and Ihh are expressed in
the kidney22,69.
1.4.2 Signaling mechanism
Patched (Ptc) is a 12 transmembrane domain receptor of the
hedgehog ligand70,71. This
tumour suppressor protein is an obligate negative regulator of
the hedgehog signaling
-
15
pathway72. Invertebrates have two patched receptors, Ptc1 and
Ptc273. They share 73%
amino acid similarity but the transmembrane domains 6 and 7 are
significantly different.
Ptc2 has been shown to have a compensatory role when Ptc1 is
lost. Unlike global Ptc1
deletion, Ptc2 loss is not embryonic lethal suggesting that Ptc1
is the primary receptor
important in embryo development and survival74.
In the absence of the hedgehog ligand, Ptc1 inhibits the
activity of another
transmembrane protein Smoothened (Smo) by inhibiting its
localization to the cilium75.
When inhibited, Smo cannot interact with an intracellular
proteolytic complex, allowing
the complex to bind and cleave full-length Gli proteins75.
Truncated Gli proteins localize
to the nucleus and act as repressors of hedgehog signaling
target genes. When
hedgehog ligand is present it binds and inhibits Ptc1, allowing
Smo to localize to the
cilium and trigger full-length Gli translocation to the nucleus
to activate gene
transcription (Figure 6)75. Dysregulation of hedgehog signaling
during embryogenesis
results in various congenital abnormalities and in adults
unrestrained hedgehog
signaling is associated with cancer development76-80.
In mammals there are three known Gli homologs: Gli1, Gli2, and
Gli381. Gli1 and Gli2
are primarily transcriptional activators and Gli3 can act as
both an activator and a
repressor. The Glis have DNA binding zinc finger domains which
can attach to a
consensus region on a target gene81. Genes activated by Gli
transcription factors
include N-myc, CyclinD1, Gli1, Gli2, and Ptc1 as well as genes
in renal patterning and
morphogenesis such as Pax2 and Sall182. Mutations in the Gli
family have been
associated with human pathologies including Greig
cephalopolysyndactyly syndrome
(GCPS) and Pallister-Hall syndrome (PHS)83,84. GCPS is an
autosomal dominant
-
16
Figure 6. Overview of hedgehog signaling. (A) In the absence of
the hedgehog
ligand transmembrane protein Patched (Ptc) inhibits Smoothened
(Smo) and Gli is
processed to a truncated form by a proteolytic complex.
Truncated Gli acts as a
transcription repressor. (B) When present, hedgehog ligand binds
Ptc abrogating the
inhibition of Smo. Active Smo prevents formation of a
proteolytic complex. Full length
Gli transcription factors translocate into the nucleus and
upregulate genes such as
Pax2, Sall1, N-myc, CyclinD1, Gli1, Gli2, and Ptc1 (adopted from
Rosenblum, 2008)23.
-
17
syndrome caused by loss of function mutations in the GLI3.
Individuals with GCPS
suffer from hypertelorism, macrocephaly with frontal bossing,
and polydactyly83. Most
commonly GCPS is caused by mutations that lead to
haploinsufficiency for GLI3,
however not all patients with GCPS have GLI3 mutations
suggesting other genes in the
GLI-hedgehog pathway could cause such a phenotype83. PHS is an
autosomal
dominant disorder characterized by multi-tissue abnormalities
including renal agenesis
or dysplasia, renal hypoplasia and hydronephrosis 85-87. PHS is
caused by frameshift
and splicing mutations of the GLI3 gene that generate a
truncated protein similar in size
to GLI3 repressor 84,88.Mutant mice that carry Gli3 mutant
alleles which terminate at
amino acid position 699 have been generated to model the frame
shift mutation and
constitutively express Gli3 repressor89. The allele is called
∆699 and it is a unique tool
for studying the role of Gli repressor and hedgehog signaling in
kidney development.
1.5 Hedgehog Signaling is Required for Kidney Development
In 2006 Hu et al demonstrated that germ line homozygous Shh
deletion causes renal
aplasia or the formation of a single ectopic dysplastic kidney
(Figure 7A)22. Analysis of
the dysplastic kidney tissue revealed decreased levels of full
length Gli1, Gli2, and Gli3
protein (transcriptional activators) an increase in the ratio of
Gli3 transcriptional
repressor to Gli3 transcriptional activator (Figure 7B)22. The
mutant kidney also
displayed a decrease in expression of known kidney patterning
genes Pax2 and Sall1
and of cell cycle genes N-myc and CyclinD122. Further
investigation of Gli3 in Shh
mutant mice revealed that homozygous deficiency of Gli3 in
Shh-/- mice rescued renal
malformations22. These data suggest that a balance between Gli
activator and Gli
-
18
Figure 7. Hedgehog signaling in kidney development. (A) Gross
anatomical features
of kidneys in wild-type and Shh–/– mice at E18.5. Shh–/– mice
exhibit either absence of
both kidneys or the presence of a single ectopic kidney located
in the pelvis. (B)
Western analysis of E14.5 kidney tissue from wild-type and
Shh–/– mice. Shh
deficiency decreases GLI1 and GLI2. GLI3 activator (190 kDa) is
also decreased, but
GLI3 repressor (89 kDa) is unaffected compared with wild type.
The ratio of GLI3
activator to GLI3 repressor is decreased from 3.25 in wild type
to 0.23 in Shh–/– mice. K
– kidney, G - gonad (adapted from Hu et al. 2006)22.
-
19
repressor proteins is required for normal kidney development and
when there is an
imbalance towards the repressor it leads to renal
maldevelopment.
In the embryonic kidney Shh is exclusively expressed in the
distal ureteric epithelium90.
During kidney development hedgehog activity is restricted to
ureter and medullary
regions but is absent from the renal cortex91. To further
delineate how hedgehog
signaling controls kidney development, ectopic hedgehog activity
in renal cortex was
examined. Rarβ2Cre is a mouse line where expression of Cre
recombinase becomes
active at E9.5 in the metanephric mesenchyme but not in the
nephric duct epithelium35.
Genetic elimination of Ptc1 using the Rarβ2Cre mouseline
resulted in hydropelvis and
thinning of the renal cortex (Figure 8AB). Another interesting
observation in mutant
kidneys was increased spatial expression of Foxd1 in the cortex
(Figure 8CD). Thus, in
the context of increased hedgehog activity in the renal cortex,
the Foxd1 positive
capsule and cortical stromal cell population increased while the
overall cortex domain
was reduced in size. These data suggested that capsule cells and
cortical stroma may
already be in a state of active hedgehog signaling unlike the
remainder of the renal
cortex. This was a novel finding because hedgehog activity was
thought to be restricted
to the ureter and medulla of a kidney.
1.6 Stromal Cells are in State of Active Hedgehog Signaling
Ptc1 LacZ reporter mice can be utilized to assess cells and
domain-specific hedgehog
activity in the embryonic kidney. Since Ptc1 is a downstream
target of hedgehog
signaling its expression is indicative of hedgehog responsive
cells (blue). Using a Ptc1
LacZ reporter, E15.5 mouse kidneys are observed to have LacZ
positive cells in the
-
20
Figure 8. Genetic elimination of Ptc1 resulting in hydropelvis
and death
immediately following birth. (A) and (B) H&E sections of
E18.5 kidneys. The Ptc1
mutants are characterized by thin renal cortex and severe
hydropelvis. (C) and (D)
Foxd1 in situ hybridization of E13.5 kidneys. The Ptc1 mutants
have increased spatial
expression of Foxd1 in the cortex (J. Cain and M. Staite
unpublished).
-
21
capsule and cortical stroma (Figure 9AB). This suggests that
there is a low level of
hedgehog activity in the capsule and cortical stroma.
Furthermore, in situ hybridization (ISH) experiment for Gli1
illustrates that the gene is
expressed in the capsule of E14.5 embryonic kidney. Gli1,
another hedgehog
downstream target gene provides further evidence that hedgehog
signaling is active in
the capsule. Lastly, the renal capsule displays prominent
expression of Ihh, a member
of the mammalian hedgehog family (Figure 9CD).
1.7 Rationale
The morphological and cellular composition defects caused by
Foxd1 deficiency
illustrate that the capsule and cortical stroma is important to
the overall development
and patterning of the kidney. Proper cortical stromal
differentiation mediated by Foxd1
defines the nephrogenic zone and its boundaries with correct
positioning of nephrogenic
units and ureteric bud tips beneath the renal capsule. However,
the molecular signals
that control formation of the capsule are largely undefined.
While hedgehog signaling is
active in the capsule, its function in this domain is unknown.
Further understanding of
hedgehog signaling in the renal capsule will provide insight
into hedgehog function in
renal capsule and capsule function during kidney
organogenesis.
1.8 Hypthesis
Hedgehog signaling is required for capsule morphogenesis in the
embryonic kidney.
1.9 Model
FoxD1Cre; Smo-/loxP mice (termed Smo null mutant) were used in
experiments.
-
22
Figure 9. Stromal cells are characterized by active hedgehog
signaling. Ptc1LacZ
expression demonstrates regions of the kidney with hedgehog
signaling activity. (A)
Strong expression of Ptc1LacZ is observed in the ureter and
medullary stroma. (B) A high
resolution image of area in (A) demarcated with a rectangle.
Weak expression is
observed in cortical stroma and capsule of the kidney. In situ
hybridization of Gli1 (C)
and Ihh (D) in E14.5 kidneys suggest hedgehog activity in the
capsule (J. Blake and
http://www.eurexpress.org/ee/)
-
23
2 . Chapter Two: Methods
2.1 Mouse Breeding
All mice used in experiments were housed at The Toronto Centre
for Phenogenomics
(TCP) animal facility (Toronto, Canada). The ethics committee at
The Hospital for Sick
Children has approved all animal experiments. FoxD1Cre mice were
mated to Smo+/-
mice to generate FoxD1Cre; Smo+/-. These mice were mated with
homozygous Smo
conditional (SmoloxP/loxP)92 mice to generate FoxD1Cre;
Smo-/loxP embryos in which Smo
was specifically removed from the renal capsule and cortical
stroma lineage in kidneys.
These kidneys are referred to as Smo null mutant kidneys.
FoxD1Cre; Smo+/- and SmoloxP/loxP mice were mated with Gli3+/-93
to generate FoxD1Cre;
Smo+/-; Gli3+/- and SmoloxP/loxP; Gli3+/- progeny respectively.
These two mouse lines were
then mated with one another to ultimately generate FoxD1Cre;
Smo-/loxP; Gli3+/- embryos
(termed Gli3 rescue mice). Polymerase chain reaction (PCR)
genotyping for each allele
was performed as previously described94. The following primers
were used for
genotyping mice. For the Smo deleted allele the forward primer
was
GGCCTGCGCTGCTCAACATGG and the reverse primer was
CCATCACGTCGAACTCCTGGC95. The forward and reverse primers for the
Smo LoxP
allele were ATGGCCGCTGGCCGCCCCGTG and GGCGCTACCGGTGGATGTGG
respectively96. Cre allele was detected using the forward
primer
GAAACAGGGGCAATGGTGCGCCTGCTG and the reverse primer
AGGAGGACGCTGGGTTGGTCCGATACT35. The Ptc1 LacZ allele was detected
using
the forward primer TGTCTGTGTGTCTCCTGAATCAC and the reverse
primer
TGGGGTGGGATTAGATAAATGCC94. The Gli3 deleted allele was detected
using 2
-
24
forward and 2 reverse primers which are able to identify
homozygote and heterozygote
embryos. Homozygotes and heterozygotes can be distinguished from
wildtype siblings
with forward primer TACCCCAGCAGGAGACTCAGATTAG and reverse
primer
AAACCCGTGCAGGACAAG that span the deletion breakpoint in
combination with a
primer set (forward: GGCCCAAACATCTACCAACACATAG and reverse:
GTTGGCTGCTGCATGAAGACTGAC) located within the deletion93.
Littermates were
used for all experiments in which wildtype (WT) and mutant
embryos were compared.
2.2 Tissue Preparation
Sacrifice of experimental animals was achieved using methods
approved by the ethics
committee at The Hospital for Sick Children (Toronto, Canada).
Pregnant mice were
sacrificed via cervical dislocation in order to isolate embryos
from timed pregnancies.
Embryonic day 0.5 (E0.5) was considered to be noon on the date
of observation of
vaginal plug. Embryonic kidneys were dissected at any time
between E11.5 and E18.5
and fixed in 4% paraformaldehyde (PFA) for at least 24 hours.
Kidneys were then
embedded in paraffin for sagittal sectioning of tissue by The
Centre for Modeling Human
Disease’s (CMHD) Core Pathology Lab at TCP using standard
methods.
2.3 β-galactosidase Staining
Whole kidneys were briefly fixed in LacZ fix solution (25%
glutaraldehyde, 100 nM
EGTA, 1 M MgCl2, 0.1M sodium phosphate) and rinsed in wash
buffer (0.1 M sodium
phosphate buffer, 2% nonidet-P40, 1M MgCl2). Kidneys were then
placed in LacZ
staining solution (25 mg/mL X-gal, potassium ferrocyanide,
potassium ferricyanide) at
37°C overnight in the dark. Once staining had occurred the
reaction was terminated in
wash buffer and post-fixed in 10% buffered formalin at 4°C.
Whole kidneys were
-
25
photographed using a Leica EZ4D dissecting microscope, processed
for embedding in
paraffin wax, and then sectioned at 5 µm. Sections were
counterstained with nuclear
fast red.
2.4 Histology and Immunofluorescence
Paraffin-embedded kidney sections were stained with hematoxylin
and eosin (HE) by
The CMHD’s Core Pathology Lab at TCP. Immunofluorescence
staining was
performed90 on formalin fixed, paraffin-embedded kidney sections
using anti-FoxD1 (a
gift provided by Dr. Andrew McMahon, 1:100 dilution), anti-lotus
tetragonolobus lectin
(LTL – Vector Laboratories, 1:100 dilution), anti-pan
cytokeratin (Sigma, 1:100 dilution),
anti-NCAM (Sigma 1:200 dilution), anti-BrDU (Roche, 1:100
dilution), and anti-Six2
(Roche, 1:200 dilution). Alexa 488 goat anti-mouse and Alexa 568
goat anti-rabbit
(Molecular Probes, 1:1000 dilution) were used as secondary
antibodies. Whole mount
immunofluorescence was performed as described with
anti-calbindin-D28K (Sigma,
1:200 dilution; secondary is Alexa 488 goat anti-mouse,
Molecular Probes, 1:100
dilution).
2.5 Capsule Cell Proliferation Analysis
Pregnant mothers were injected with Bromodeoxyuridine (BrDU)
(100 mg/Kg) and
tissue of E11.5 and E13.5 embryos was harvested 2 hours after
injection. Harvested
kidneys were stained with a fluorescent anti-BrDU antibody to
identify cells that
incorporated BrDU. Since mutant capsule cells could not be
counterstained with a
fluorescent FoxD1 antibody, capsule cells were identified,
disticnt from mesenchymal
cells by their morphology. In order to count every capsule cell
in the kidney several
criteria were established to aid the counting process. These
criteria were: 1) capsule
-
26
cells are on the outer edge of the kidney, 2) capsule cells have
a flattened appearance,
3) capsule cells are a maximum 3 cell layers away from the edge
of the kidney, and 4)
capsule cells are not marked by a fluorescent Six2 marker (some
sections were
counterstained with a Six2 antibody). Using these 4 criteria and
the fluorescent BrDU
stain a ratio of proliferative capsule cells over total number
of capsule cells was
calculated.
2.6 In situ mRNA Hybridization
Whole mount embryos were fixed in 4% PFA in PBS for 16 hours at
4°C. In situ
hybridization was performed on paraffin-embedded sections using
DIG-labeled cDNA
probes encoding Bmp4, Foxd1, Gdnf, Gli1, Ptc1, Raldh2, ret,
Wnt4, and Wnt11 as
previously described97.
2.7 Kidney Volume Analysis and Total Nephron Number Count
Whole E18.5 harvested kidneys were fixed in 4% PFA in PBS for 16
hours at 4°C. The
kidneys were paraffin embedded and sagittaly sectioned by the
CMHD’s Core
Pathology Lab at TCP according to the following instructions:
the first 100 µm of tissue
was discarded, then a slide with 4 sections, each 5 µm thick,
was prepared and the next
80 µm were discarded again98. These steps were repeated
throughout the entire
volume of each kidney. All the sections were then stained with
Periodic acid-Schiff
(PAS) stain. The best section of each slide was imaged and using
Adobe Photoshop
CS5.1 the surface area (SA) was calculated. The kidney volume
was calculated by
multiplying 100 µm to each section’s SA value and adding up all
the individual volumes.
-
27
The PAS stain highlights the basement membranes of glomerular
capillary loops, thus
the glomeruli were very visible in each section. The best
section of each slide was again
utilized to tally all mature glomeruli with a visible Bowman’s
space. The sum of all the
sections making up the kidney was used as the total nephron
number per kidney.
2.8 Scanning Electron Microscopy
Whole E16.5 kidneys and E13.5 embryos were harvested and fixed
in 2%
glutaraldehyde in 0.1M sodium cacodylate buffer pH 7.3 overnight
at 4°C. The samples
were given to The Advanced Bioimaging Centre at Mount Sinai
Hospital (Toronto,
Canada) to be dehydrated, dried, and gold sputter-coated. The
prepared samples were
imaged at the Bioimaging Centre with a FEI XL30 ESEM
microscope.
2.9 Blood and Urine Collection
Blood and urine collection was performed according to the
protocol provided by CMHD.
Conscious mice were restrained in an uncapped 50 mL Falcon tube
with air holes drilled
in the closed end. The left or the right leg was extended and
fixed firmly by holding the
fold of the skin between the tail and the thigh. Enough hair was
then removed from the
surface of the fixed leg to expose and visualize the saphenous
vein. The exposed skin
was then wiped clean with 70% ethanol and dried with a piece of
gauze. A small
amount of Vaseline was applied on the shaved skin to reduce
clotting and help prevent
the blood from collecting in the remaining hair on the leg.
Using a 25 gauge needle the
vein was punctured. As a drop of blood appeared on the surface
of the leg a capillary
tube held on a 45° angle was used to collect the blood and
dispense into a 0.5 mL
microtube. The tube was capped and mixed by flicking the side of
the tube and stored
on crushed ice for biochemistry analysis. Approximately 100 µL
of blood was collected
-
28
from each mouse. Once enough blood was collected the mouse’s
foot was flexed to
reduce flow of blood to the puncture site and a gauze compress
was applied to stop
bleeding. The blood biochemistry analysis was performed by the
CMHD Mouse
Physiology Facility.
Mouse urine was collected in 1.5 mL microtubes from conscious,
restrained mice. The
mouse was picked up and held by the excess skin at the base of
the neck. The
microtube was held at the point of urination and the bladder of
the mouse was gently
massaged. Any expelled urine was collected in the microtube. The
urine in the tube was
thoroughly mixed and given to the CMHD Mouse Physiology Facility
for urinalysis.
Urinalysis was performed using Chemstrip 4MD test strips (Roche
Diagnostics).
2.10 Data Analysis
Statistical analysis was performed using GraphPad Prism software
(version 5.0). Data
were analysed using a unpaired Student’s t-test. A probability
of less than 0.05 was
considered to indicate statistical significance.
3 . Chapter Three: Results
3.1 Smo null mutant kidneys are characterized by decreased
expression of HH target gene mRNA
To begin to investigate the functions of the hedgehog signaling
pathway in capsule and
cortical stromal cells, mice were generated with deficiency of
Smo specifically using
Cre-recombinase directed by a Foxd1 promoter element. These
kidneys are referred to
as Smo null mutant kidneys. If Foxd1Cre is efficient and Smo is
successfully deleted in
the capsule then hedgehog activity should be diminished in this
structure. mRNA levels
of Ptc1 and Gli1 were used as reporters of hedgehog
activity.
-
29
E13.5 embryos were sectioned and ISH was performed to study the
expression of Gli1
and Ptc1. In wildtype embryos Gli1 expression is widespread both
in the cortex and
parts of the medulla of kidneys (Figure 10ABC). In the Smo null
mutant, Gli1 expression
was lost in the capsular region but was maintained in the
remainder of kidney tissue
(Figure 10DEF). Similarly, ISH revealed that Ptc1 expression is
completely lost in the
Smo null mutant capsule whereas the wildtype mouse displayed
faint expression in the
same region (Figure 11). Together, these data demonstrate that
Smo deletion was
specific to the capsule and that hedgehog activity was
deactivated only in that region.
3.2 Patterning of the outer renal cortex is abnormal in Smo null
mutant kidney
To investigate the phenotype of the Smo null mutant, whole P0
kidneys from wildtype
and Smo null mutant littermates were collected and fixed in PFA.
Following sectioning
of kidneys and staining with HE, the Smo null kidneys were
analyzed by microscopy.
They were observed to have unusual patterning and were slightly
smaller. In wildtype
kidneys there is a distinct renal capsule, comprised of a layer
of flattened cells that are
just above the cortex, which contains the nephrogenic zone
(Figure 12AC). The outer
cortex made up of the nephrogenic zone contains condensing
mesenchyme with
nephron precursors while the inner cortex is composed of older
generations of mature
nephrons and tubule structures. The Smo null mutant kidneys
display different
morphology. The capsule around the Smo null kidneys does not
surround the entire
kidney. There are regions on the outer edge of the kidney where
no capsular cells are
present. Furthermore, the cortex underneath these regions is
made up of mature
glomeruli and tubules rather than a nephrogenic zone (Figure
12BD).
-
30
Figure 10. In situ hybridization of Gli1 at E13.5. (A) In
wildtype kidneys Gli1
expression is observed in both the cortex and medulla. (B) In
the medulla Gli1 is
strongly expressed in the ureter (U) of a wildtype kidney. (C)
In the cortex, Gli1
expression is widespread and can be observed in the capsule (C)
as well as some
ureteric tips and mesenchymal cells. (D) The domain of Gli1
expression in kidneys of
FoxD1Cre; Smo-/loxP (Smo null) mice was similar to wildtype mice
except the capsule.
(E) In the ureter Gli1 expression was maintained. (F) Gli1
expression was not detected
in the capsule, however other cortical structures that displayed
Gli1 expression in
wildtype mice maintained it in the mutant. U – ureteric branch,
C – capsule.
-
31
Figure 11. In situ hybridization of Ptc1 at E13.5. (A) Weak Ptc1
expression is
detected in the capsule (C) of wildtype kidneys.(B) In Smo null
kidney, Ptc1 expression
was absent in the capsule, however expression was still present
in the ureter (U) of
mutant kidneys. C – capsule, U – ureter.
-
32
Figure 12. Hematoxylin and eosin (H&E) stained P0 kidneys.
(A) The wildtype
kidney has an outer cortex (O.C. consists of the capsule,
cortical stroma and the
nephrogenic zone), an inner cortex (I.C. consists of tubules and
mature glomeruli) and a
medulla (M consists of ureteric branches and medullary stroma)
(C) Wildtype kidney
has a distinct renal capsule (C) and well defined a nephrogenic
zone (NZ). (B) and (D)
The Smo null kidney appears to have a discontinuous capsule and
the nephrogenic
zone is disrupted by tubules and mature glomeruli. I.C. – inner
cortex, O.C. – outer
cortex, M – medulla, C – capsule, NZ – nephrogenic zone.
-
33
To define the structure of the capsule scanning electron
microscopy (SEM) was utilized
to observe the surface of kidneys. E16.5 kidneys of wildtype and
Smo null embryos
were dissected and fixed in 2.5% glutaraldehyde and sent to
Advanced Bioimaging
Centre at Mount Sinai Hospital. SEM imaging illustrated that the
capsule cells on the
surface of the wildtype kidney surround the entire organ and
endow the kidney a flat
and smooth appearance (Figure 13AD). In contrast the Smo null
mutant kidney at this
time point displayed a discontinuous layer of capsule cells
(Figure 13BCE). The Smo
null mutant kidneys exhibited various regions on the surface
where capsule cells were
absent and instead round mesenchymal cells were visible giving
the kidney a rough
appearance.
3.3 Smo null mutant capsule appears discontinuous after
E13.5
In order to ascertain at what stage in kidney development the
capsule of the Smo null
mutants becomes discontinuous histological sectioning was
utilized. Time points from
the latest (E18.5) to earlier developmental stages were examined
until a time point was
discovered where the capsule of the mutant did not appear
discontinuous. HE sections
of wildtype and mutant kidneys at E18.5, E15.5 and E13.5 were
imaged and analyzed.
The images revealed that mutant kidneys at E18.5 and E15.5
definitively displayed the
discontinuous capsule phenotype. However, HE sections of E13.5
wildtype and mutant
kidneys had intact capsule and cortical stroma which were
indistinguishable from one
another (Figure 14). This result suggested that the
discontinuous capsule phenotype
presents after E13.5. This was further confirmed with SEM
imaging which showed that
at E13.5 flat capsule cells on the surface of both wildtype and
Smo null mutant kidneys
surrounded the entire organ (Figure 15).
-
34
Figure 13. Scanning Electron Microscopy (SEM) images of E16.5
wildtype and
Smo null kidneys. (A) and (D) A flattened layer of capsule cells
enveloping the
wildtype kidney gives it a smooth appearance. (B), (C), and (E)
The mutant kidney
appears to have a segmented capsule with the regions that have
exposed cells that
normally underlie the capsule.
-
35
Figure 14. H&E sections of wildtype and Smo null kidneys at
various embryonic
stages. The discontinuous capsule phenotype is observed in
mutant kidneys of E18.5
(F) and E15.5 (E) embryos. (A) and (D) However at E13.5 the
kidneys look fairly similar.
C – capsule.
-
36
Figure 15. SEM images of E13.5 wildtype and Smo null kidneys.
Capsule cells on
the surface of both wildtype (A) and Smo null kidneys (B)
surrounded the entire kidney.
A flattened layer of cells enveloping the kidney gives it a
smooth appearance.
-
37
3.4 Smo null mutant kidneys are smaller in volume and have fewer
nephrons than wildtype kidneys
Kidney volume was estimated to establish if there was a
significant change in kidney
size. Smaller kidneys also may indicate lower number of
nephrons, therefore the total
number of glomeruli was quantified as well.
Kidneys of E18.5 wildtype and mutant littermates were serially
sectioned every 100 µm
(diameter of a mouse glomerulus is approximately 100 µm). The
surface area of each
section was determined using an analytical tool in Adobe
Photoshop CS5.1. By
multiplying each surface area value by 100 µm the total volume
of the kidney was
estimated. The average volume of 8 wildtype and 8 mutant kidneys
were found to be
2.97 mm3 and 1.53 mm3 respectively (Figure 16A). There was
approximately a 48%
decrease in volume of Smo null kidneys. The count of glomeruli
was performed on 6
wildtype and 6 mutant kidneys. Glomeruli tallied in the count
had to be surrounded by
Bowman’s space and have visible basement membranes, as
highlighted by the PAS
stain. The final count resulted in an average of 238 glomeruli
in wildtype kidneys and
139 in mutants, signifying a 42% decrease (Figure 16B). While
the mutant kidneys had
fewer glomeruli, the morphology of those that were present was
the same as those
found in wildtype kidneys (Figure 16CD).
3.5 Smo null mutant kidneys have decreased Foxd1 and Raldh2
expression
The loss of Smo in capsule cells results in abnormalities in the
morphology of the
capsule, thus further investigation was done to explore if the
defect correlates with
alterations in the cellular composition and activity of the
capsule.
-
38
Figure 16. Kidney volume and nephron number analysis of E18.5
kidneys. (A) The
average volume of 8 wildtype and 8 mutant kidneys were found to
be 2.97 mm3 and
1.53 mm3 respectively. There was approximately a 48% decrease in
volume of Smo null
kidneys. (B) The average total number of nephrons counted in
control and Smo null
kidneys were 238 and 139 respectively, a 42% decrease in the
mutant. (C) and (D)
Morphology of glomeruli in wildtype and mutant kidneys are
similar.
C D
-
39
The expression of renal capsule molecular markers, Foxd1 and
Raldh2, examined by
ISH at time points after the Cre recombinase becomes active. At
E11.5 in wildtype
kidneys Foxd1 expression was limited to the outer edge of the
kidney where the capsule
progenitors are situated. Conversely, Foxd1 expression in Smo
null mutant kidneys was
still present in the same region but diminished compared to
wildtype when examined by
ISH. The decrease in Foxd1 expression was persistent and more
pronounced by E13.5.
To demonstrate that Foxd1 protein levels correlates to
expression detected by ISH, an
immunofluorescence stain of Foxd1 at E15.5 did not detect the
protein in mutant
kidneys (Figure 17EF). This data indicates that Foxd1 expression
is perturbed and
decreased in a model of decreased hedgehog signaling in Foxd1
progenitors.
A similar result was observed with Raldh2 expression. At E13.5
Raldh2 expression was
limited to the capsule of the kidney and the cortical stroma
which slightly penetrates into
the cortex. In the Smo null mutant Raldh2 expression was present
in the same regions
as the wildtype kidney, however the signal was not as intense.
Raldh2 expression was
almost undetected in E15.5 mutant kidneys, while remained robust
in wildtype (Figure
18).
Together these data show that molecular markers normally
expressed in the capsule
are not detected at the same levels throughout development in
the mutant kidney. This
suggests that hedgehog activity may be required for either
establishing or maintaining
normal cellular activity in the capsule to allow proper
maturation.
-
40
Figure 17. In situ and immunofluorescence of Foxd1 at various
stages. Foxd1 is
expressed on the outer edge of E11.5 (A), E13.5 (C) and E15.5
(E) wildtype kidneys.
Conversely, Foxd1 expression in Smo null kidneys is still
present in the same region but
appears diminished compared to wildtype at E11.5 (B) and E13.5
(D). At E15.5, no
Foxd1 protein is observed (F).
-
41
Figure 18. In situ hybridization of Raldh2. Raldh2 is expressed
in the capsule and
cortical stromal cells of E13.5 (A) and E15.5 (C) wildtype
kidneys. Conversely, Raldh2
expression in the E13.5 Smo null kidneys (B) was greatly
diminished compared to
wildtype and in the E15.5 Smo null kidney was hardly detected
(D).
-
42
3.6 Proliferative capacity of mutant capsule cells is
significantly decreased
The Smo null mutant capsule looks fairly normal at early
embryonic stages but appears
increasingly segmented throughout development. Does this occur
because capsular
cells at an early embryonic age do not proliferate as well,
resulting in not enough cells to
be able to surround a larger kidney?
To investigate this question, proliferation was quantified in
the renal capsule by BrDU
injection of pregnant mothers for two hours at E11.5 and E13.5.
Harvested kidneys
were stained with a fluorescent anti-BrDU antibody to visualize
cells that were dividing.
Utilizing the identifiable morphology of capsule cells a ratio
of proliferative capsule cells
over total number of capsule cells was calculated.
At E11.5 the ratios of wildtype and mutant kidneys were 0.275
and 0.222 respectively.
There was a significant 20% decrease in the proliferative
capacity of mutant capsule
cells. The effect was more prominent by E13.5 when the ratios of
wildtype and mutant
kidneys were 0.268 and 0.124 which corresponded to a 54%
decrease in proliferative
capacity (Figure 19).
This experiment showed that Smo deletion in capsular stroma
causes a decrease in the
rate of proliferation of capsule cells. Discontinuity of the
mutant capsule is likely a result
of the loss of proliferative capacity in capsule cells.
-
43
Figure 19. A ratio of BrDU positive (proliferative) capsule
cells over total number
of capsule cells was used as an index of proliferative capacity.
(A) Kidney stained
with a Six2 (red) and BrDU (green) antibody and DAPI (blue). All
capsule cells (C) were
separately counted from mesenchymal cells (M) by morphology and
location. The white
segmented line represents the outer edge of the kidney where
capsule cells are located.
(B) At E11.5 the proliferative index of wildtype and mutant
kidneys were 0.275 and
0.222 respectively. There was a 20% statistically significant
(P
-
44
3.7 Mutant mice do not always have a discontinuous capsule
The deletion of Smo in Foxd1 progenitors has a significant
impact on the morphology of
the kidney. Are mutants viable and if so does this phenotype
affect kidney function?
Pregnant mice were permitted to undergo a full term pregnancy
and give birth to their
pups to examine the viability of mutants. The first litter
examined was immediately after
birth.
A litter of six P0 pups was harvested of which three were
discovered to be deceased.
Genotyping analysis revealed that all three deceased pups were
mutants, while the
three viable pups were a mix of mutant and wildtype alleles. HE
sections of kidneys
from wildtype, viable mutant, and non-viable mutant kidneys
revealed that viable mutant
kidneys looked similar in structure to wildtype kidneys. Viable
mutants had an intact
capsule which surrounded the whole renal organ. Non-viable
mutant kidneys displayed
the discontinuous capsule phenotype and had mature nephrons near
the surface as
described previously (Figure 20).
These data indicate that deletion of Smo in Foxd1 progenitors
has a variable phenotypic
outcome. Some mutants have a continuous capsule and an ample
number of cortical
stroma cells surrounding the mesenchyme regardless of the Smo
deletion. Further,
investigation needs to be done to determine the proliferative
index of capsule cells of
mutant viable mice. If mutant mice have a normal capsule, there
may not be a
significant decrease in the rate of proliferation of capsule
cells in those mice. Future
evidence found to support this hypothesis will imply the
presence of another signal that
stimulates cell division in the absence of hedgehog
signaling.
-
45
Figure 20. H&E sections of wildtype, viable mutant, and
non-viable mutant P0
kidneys. Viable mutant kidneys (BE) look similar in structure to
wildtype kidneys (AD).
Viable mutants have an intact capsule which surrounds the whole
renal organ. Non-
viable mutant kidneys (CF) display the discontinuous capsule
phenotype and have
mature nephrons near the surface.
-
46
3.8 Viable mutants display normal kidney function
Urinalysis and blood biochemistry was performed on adult mutant
mice to ascertain if
Smo deletion has an effect on kidney function. To obtain an
index of renal function,
there are a number of quick and non-invasive or minimally
invasive measures that can
be performed.
Two of the most fundamental indicators of renal function that
can be measured from a
mouse blood sample are plasma creatinine and blood urea nitrogen
(BUN). The body’s
excretion of creatinine and BUN is a primarily dependent on
glomerular filtration,
therefore an elevation of either of these in the blood indicates
reduced renal filtration
capacity and may point to an underlying renal disease. Another
marker of renal function
is uric acid as 70% of its disposal occurs via the kidneys, and
in 5-25% of humans,
impaired renal excretion leads to hyperuricemia99. Blood levels
of creatinine, urea, and
uric acid in mutant mice were not significantly different from
their wildtype littermates
suggesting normal filtration capacity (Figure 21).
Another non-invasive method for assessing renal function is to
measure albumin or total
protein excretion in the urine. Healthy kidneys excrete minimal
amounts of protein, thus
proteinuria is commonly used as a sign of kidney damage. The
protein levels in the
urine of all mice were normal further indicating no change in
kidney function. Additional
testing did not detect any blood in all urine samples and
glucose concentration was also
normal in almost all samples (Table 1).
-
47
Figure 21. Blood biochemistry of adult wildtype and Smo null
littermate mice.
Blood levels of creatinine, urea, and uric acid in mutant mice
were not significantly
different from their wildtype littermates suggesting normal
filtration capacity (P>0.05 for
all metabolites).
-
48
Table 1: Urinalysis of adult wildtype and Smo null littermate
mice.
-
49
3.9 Gli3 deficiency improves renal patterning and capsule
formation
The work of Hu et al, 200622 demonstrated that the ratio of Gli3
activator and repressor
is very important to proper development of the kidney.
Furthermore, at low levels of
hedgehog signaling, as is the case in the renal capsule, Gli3
does not get cleaved to a
repressor form and the repressor targets are expressed. However,
hedgehog signaling
is absent in the capsule of the Smo null mutants which may
result in Gli3 being
proteolytically processed to the repressor form. To investigate
Gli3 activity in capsule
development, Gli3 alleles were removed to preclude formation of
the Gli3 repressor.
Analysis of HE stained sections showed that with the loss of
even one Gli3 allele the
patterning of Smo null mutant kidney structures were restored
and appeared to be
similar to wildtype(Figure 22 and 23). The capsule of Smo null
mutants with one or both
Gli3 alleles deleted was no longer discontinuous and enveloped
the entire kidney
surface. However, there was a lack of cortical stroma present
beneath the capsule
which surrounds and supports condensing mesenchyme resulting in
the surface of
kidneys not having a flat and smooth appearance as in wildtype
kidneys (Figure 22).
When both Gli3 alleles are deleted, kidneys have smoother
appearance as cortical
stroma cells are present in regions between condensing
mesenchyme. The capsule
layer while present in both Gli3 mutants appears thinner when
compared to wildtype
kidneys.
4 . Chapter Four: Discussion and Future Experiments
The renal capsule is an important structure not only in shaping
the kidney but also in the
overall normal patterning of the kidney. Current evidence
suggests that a developing
kidney has a delicate balance of signals which are imparted from
the capsule and
-
50
Figure 22. H&E sections of wildtype, Smo null mutant, and
Smo null; Gli3
heterozygous deleted mutant P0 kidneys. (A) and (D) The wildtype
kidney has a
distinct renal capsule and well defined a nephrogenic zone as
expected. The arrows in
(D) point out cortical stroma which surround and support cap
mesenchyme. (B) and (E)
The Smo null kidney has a discontinuous capsule and disrupted
nephrogenic zone as
previously described. (C) and (F) A FoxD1Cre; Smo-/loxP;Gli3+/-
(Smo null; Gli3 het)
mutant has normal tissue patterning comparable to the control
kidney. The nephrogenic
zone is not disrupted and the capsule layer surrounds the entire
kidney. Arrows indicate
regions where more cortical stroma should be to present to
surround cap mesenchyme.
The surface of the kidney does not appear smooth like the
wildtype.
-
51
Figure 23. H&E sections of control, Smo null mutant, and Smo
null; Gli3 null
mutant E18.5 kidneys. (A) and (D) The control kidney is
FoxD1Cre; Smo+/loxP. No
wildtype embryos were harvested in this litter. With the loss of
one Smo allele in the
capsule the kidney still has an intact renal capsule and well
defined a nephrogenic
zone. The capsule layer is thinner than what is normally
observed in wildtype kidneys.
(B) and (E) The Smo null kidney has a discontinuous capsule and
disrupted
nephrogenic zone as previously described. (C) and (F) A
FoxD1Cre; Smo-/loxP;Gli3-/-
(Smo null; Gli3 null) mutant has normal tissue patterning
comparable to the control.
Furthermore, the nephrogenic zone is not disrupted and the
capsule layer surrounds the
entire kidney. The capsule layer is thin like in the
control.
-
52
cortical stroma and from the interior to pattern the nephrogenic
and ureteric
compartments of the nephrogenic zone51. Taken together these
signals allow for
controlled radial expansion of the kidney and maintain proper
compartmentalization.
4.1 Deleterious effects on capsule formation by decreased
hedgehog signaling
Hedgehog signaling has never been previously implicated in renal
capsule
development. The Smo null mutant mice had a significant
phenotype which is never
observed in any wildtype mice or mice with the Foxd1Cre allele,
highlighting the
importance of the signaling pathway in capsule formation.
Foxd1Cre mediated deletion
of Smo renders capsule cells in an inactive state regardless of
the presence of a
hedgehog ligand. The specific loss of expression of Ptc1 and
Gli1, genes normally
upregulated by hedgehog signaling, in the capsule area strongly
suggest that these
mice are a good model to study capsule development in the
context of inactive
hedgehog signaling. Smo null mutant kidney capsule cells do not
proliferate as
frequently, consistent with previous knowledge that the hedgehog
pathway activates cell
cycling genes N-myc and CyclinD1. The mouse metanephros at E11.5
is composed of
approximately 1000 mesenchymal cells which go on to
differentiate and give rise to a
large number of nephrons (~13000) and other structures that
comprise the kidney1.
There is considerable change in size and volume that is
associated with the
development of the kidney from 1000 mesenchymal cells to a
mature organ. The
capsule cells must continuously proliferate and maintain a high
cell population in order
to be able to surround kidney as it becomes larger. Loss of
hedgehog signaling in these
cells has a significant impact on proliferative capacity as
measured by BrDU
incorporation. While the cells are still able to divide, the
rate of proliferation is
-
53
considerably lower and is outmatched by the rate of increase in
kidney volume. The
result is a discontinuous capsule phenotype where regions of the
surface of the kidney
are not coated with capsule cells due to a lack of
availability.
In addition to loss of proliferative capacity, deactivated
hedgehog signaling also resulted
in gradual decrease of normal capsule markers Foxd1 and Raldh2.
Foxd1 expression
was not immediately lost, therefore the resulting phenotype was
not the same as what
was observed in Foxd1 null mice from Levinson et al, 2005. Foxd1
null mice have
kidneys which fuse to one another at the midline and do not
detach from the dorsal
body wall47. While levels of Foxd1 were diminished as early as
E11.5 in the Smo null
mutant, the outcome was not similar to the Foxd1 mutant, kidneys
separated from one
another and the surrounding body wall by E15.5. Similar to the
Foxd1 null mice as
Foxd1 expression diminished, Raldh2 was also lost. This data
suggests hedgehog
signaling is at least indirectly responsible for maintaining
expression of these capsule
markers. Further experiments exploring hedgehog inactivity at
earlier time points will
determine whether hedgehog signaling is required to initiate
Foxd1 and Raldh2
expression as well.
4.2 Capsule formation is required for viability
Capsule formation is not only important for kidney development,
but also for embryo
viability as highlighted by the Smo null mutant mice not being
viable. However, there is
a caveat that if embryos with Smo deficiency can form a capsule
that envelops the
entire kidney, they become viable and grow to adulthood with
normal kidney function.
How can they form a full capsule layer? It is possible that the
viable mice are
characterized by incomplete excision of Smo resulting in only
partial loss of hedgehog
-
54
signaling. The BrDU injections demonstrate that capsule cells
continue to proliferate;
only the rate is decreased. Partial hedgehog activity may be
able to maintain the
proliferative index of capsule cells so that the capsule does
not become discontinuous.
If excision of Smo is complete and hedgehog signaling is
inactive, there must be other
active pathways that stimulate capsule cells proliferation. It
is possible that in viable
mutants, other pathways get upregulated to compensate for the
deactivated hedgehog
pathway. If an alternative pathway can promote capsule cell
proliferation, the
upregulation of that pathway may be inconsistent. This may be
evidenced by the
observation that non-viable mutants with a discontinuous capsule
display variability in
the severity of the phenotype. Specifically, some kidneys have
exposed regions in the
capsule every 100 µm around the entire organ while less severely
affected kidneys
have few sections on the surface where the capsule is
discontinuous. Weak or strong
upregulation of an alternate pathway may result in a less
severely affected non-viable
mutant or an unaffected viable mutant respectively. Surprisingly
the presence of the
capsule is sufficient to correct for other renal patterning
issues and kidney function
remains unaffected. This may suggest that signals which are
imparted from the capsule
and help establish the nephrogenic zone are independent of
hedgehog signaling.
4.3 Effects of loss of capsule on renal patterning
Although, expression of capsule markers of E13.5 wildtype and
Smo null mutant
kidneys is dissimilar, histological sections of the two are
indistinguishable. Past E13.5,
the capsule of mutant kidneys becomes discontinuous exposing
areas on the surface
that would otherwise lie beneath the capsule. Patterning within
Smo null mutant kidneys
looks dysplastic as the capsule becomes discontinuous. Exposed
regions, no longer
-
55
covered by capsule cells, are not patterned like the remainder
of the outer cortex.
Components of the outer cortex, including cortical stroma,
ureteric bud tips, and
aggregates of nephron progenitors, are not observed in exposed
regions of mutant
kidneys. Structures which compose the inner cortex such as
mature glomeruli and
tubules are instead present on the surface.
Although the deletion of Smo is specific to capsule cells, a
deleterious effect on the
nephrogenic zone where capsule cells are lacking is also
observed. The corresponding
loss of the nephrogenic zone to the loss of capsule cells
suggests that capsule cells
play an important role in nephrogenic zone maintenance. In
wildtype mice, glomerular
units are not observed near the surface where the nephrogenic
zone is. The presence
of morphologically normal glomerular units near the surface
where capsule cells are
absent suggests that signaling molecules that differentiate
nephrogenic units are
maintained while signals that control their localization are
absent. Therefore, capsule
cells may generate signals which define and maintain the
nephrogenic zone preserve
the boundary between nephrogenic and medullary zones and their
respective cell types.
Previous work by Yallowitz et al, 201149 demonstrated that
continued expression of key
nephrogenic mesenchymal markers is dependent on proper
integration of Foxd1-
expressing capsule cells around metanephric mesenchyme.
The balance between the capsule layer, the nephrogenic zone, and
the medullary zone
is perturbed when capsule cells do not surround condensing
mesenchyme. The
importance of capsule cells signaling in renal development is
distinctly apparent when it
is absent and controlled radial expansion of kidneys gets
disrupted. The disturbance of
-
56
signals that establish the nephrogenic zone is most likely the
largest contributor to the
large decrease in kidney volume and nephron number observed in
the mutants.
4.4 Gli repressor influences renal capsule development
In Smo null mutant mice hedgehog signaling is inactivated in
capsule cells. It is difficult
to interpret in this model whether the phenotypes associated
with this mutant are
caused by too much repressor or too little activator in the
capsule. In order to consider
these possibilities, compound mutants were generated that had
decreased or no Gli
repressor. The results illustrated that Gli repressor has a
significant role in capsule
formation. The capsule was not discontinuous and patterning of
structures within the
cortex and medulla was normal in Gli3 rescue mice even though
the kidneys remained
hypoplastic. However, Gli3 deficient kidneys are known to be
reduced in volume by
15%91. This data suggests that active hedgehog signaling in
capsule cells prevents Gli
repressor activity. Furthermore, this data does not rule out the
possibility that insufficient
Gli activators are