SUPPLEMENTAL DATA Supplemental Protocol S1. Laser capture microdissection, RNA isolation and amplification. Butyl methyl methacrylate (BMM) was used as an embedding medium for structural preservation of ovule tissues as specified in Materials and Methods. This medium has been used routinely in cytological analyses, particularly for in situ hybridization in sexual and apomictic Hieracium. The BMM matrix is readily removed with acetone prior to laser microdissection (Baskin et al., 1992; Tucker et al., 2003; Okada et al., 2007; Tucker et al., 2012a). Histological detail was well preserved after acetone treatment and individual cells such as AI cells and other tissue layers remained relatively easy to identify prior to LCM (Figure S1A). However, when the Leica laser was used under the conditions described in Materials and Methods to dissect cell types it left a broad trace (Figure S1B-C) where adjoining cells were considered to be destroyed as per other published experiments using LCM. This trace is comparable to that previously reported in other studies dissecting larger tissue masses (Day et al., 2005) but not as fine as the uniquely modified laser used by Wuest et al. (2010) to dissect individual Arabidopsis mature female gametophyte cells. We conducted experiments to examine if we were able to isolate RNA from distinct cell types in sufficient quantity for 454 sequencing analyses that would enable detection of low copy genes. In preliminary LCM experiments, we examined the recovery of RNA when ovules were excised from 5μm thick ovary sections by laser capture (Figure S1A; white dashed line). We estimated that each captured ovule section contained 250 “cells”. When 20 ovule sections containing approximately 5,000 “cells” were captured and the RNA extracted using a PicoPure kit (Arcturus Bioscience Inc, Mountain View, CA, USA), the RNA recovery was low and difficult to quantify (NanoDrop spectrophotometer, Thermo Scientific, Wilmington, DE, USA). Therefore, the quantity of RNA was estimated by indirect measurement using RT- PCR on known quantities of whole ovary RNA samples to detect the presence of a low level ovary expressed gene MAP3K (ID 7.01 Table S2). The intensity of a band generated from using RNA extracted from 20, and 2, captured ovule sections was compared to that generated from a PCR reaction with 25ng, 5 ng and 1 ng of whole ovary input RNA (Figure S1D). From this we estimated that the RNA isolated from 20 ovule sections was approximately 1–2 ng. As this equates to approximately 5,000 “cells”, we considered it was unrealistic to harvest
24
Embed
SUPPLEMENTAL DATA Supplemental Protocol S1.€¦ · capture microdissected cell types from H. praealtum ovule sections. A, An ovary longitudinal section showing an ovule outlined
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
SUPPLEMENTAL DATA
Supplemental Protocol S1. Laser capture microdissection, RNA isolation and amplification.
Butyl methyl methacrylate (BMM) was used as an embedding medium for structural
preservation of ovule tissues as specified in Materials and Methods. This medium has been used
routinely in cytological analyses, particularly for in situ hybridization in sexual and apomictic
Hieracium. The BMM matrix is readily removed with acetone prior to laser microdissection
(Baskin et al., 1992; Tucker et al., 2003; Okada et al., 2007; Tucker et al., 2012a).
Histological detail was well preserved after acetone treatment and individual cells such as AI
cells and other tissue layers remained relatively easy to identify prior to LCM (Figure S1A).
However, when the Leica laser was used under the conditions described in Materials and
Methods to dissect cell types it left a broad trace (Figure S1B-C) where adjoining cells were
considered to be destroyed as per other published experiments using LCM. This trace is
comparable to that previously reported in other studies dissecting larger tissue masses (Day et
al., 2005) but not as fine as the uniquely modified laser used by Wuest et al. (2010) to dissect
individual Arabidopsis mature female gametophyte cells. We conducted experiments to
examine if we were able to isolate RNA from distinct cell types in sufficient quantity for 454
sequencing analyses that would enable detection of low copy genes.
In preliminary LCM experiments, we examined the recovery of RNA when ovules were
excised from 5µm thick ovary sections by laser capture (Figure S1A; white dashed line). We
estimated that each captured ovule section contained 250 “cells”. When 20 ovule sections
containing approximately 5,000 “cells” were captured and the RNA extracted using a PicoPure
kit (Arcturus Bioscience Inc, Mountain View, CA, USA), the RNA recovery was low and
difficult to quantify (NanoDrop spectrophotometer, Thermo Scientific, Wilmington, DE,
USA). Therefore, the quantity of RNA was estimated by indirect measurement using RT-
PCR on known quantities of whole ovary RNA samples to detect the presence of a low level
ovary expressed gene MAP3K (ID 7.01 Table S2). The intensity of a band generated from
using RNA extracted from 20, and 2, captured ovule sections was compared to that generated
from a PCR reaction with 25ng, 5 ng and 1 ng of whole ovary input RNA (Figure S1D).
From this we estimated that the RNA isolated from 20 ovule sections was approximately 1–2
ng. As this equates to approximately 5,000 “cells”, we considered it was unrealistic to harvest
AI “cells” under these experimental conditions and obtain sufficient RNA for transcriptome
generation without amplification
In the next series of experiments, RNA extracted from 2 captured ovule sections ~500
“cells”) predicted to contain 0.1-0.2 ng RNA was amplified once using the MessageAmpTM
II
aRNA Kit (Ambion Inc, Austin, Texas, USA). Indirect measurement by RT-PCR using 2 μl
out of the total 100 μl of aRNA was equivalent to the 5 ng RNA input amplification control
indicating a total yield of 250 ng aRNA. We concluded that harvesting “100–250” individual AI
“cells” from 5 µm sections by laser capture would be sufficient material for RNA isolation,
amplification and use in downstream transcriptome experiments, particularly if a second
amplification step was included. Figure S1F shows that the amplified RNA was free of
genomic DNA contamination and Figure S1G shows that the Hieracium FIE gene was detected
in all samples as expected, while the HDMC1 gene expressed during meiosis in megaspore
mother cells was absent in the three samples. Remaining experiments focused on accumulating
captured cell samples and isolating RNA, amplification as described in Materials and Methods
and quantifying yields (Table S2).
Supplemental Figure S1. Validation of quality and quantity of amplified RNA from laser
capture microdissected cell types from H. praealtum ovule sections.
A, An ovary longitudinal section showing an ovule outlined in white and an AI cell
outlined in yellow before laser microdissection. B, The same section in A after
tracing around the AI cell with the laser. The pink shading indicates the damage path
of the laser. The intact AI cell is outlined in yellow. C, View of the section in B after
the AI cell was harvested into the cap of the capture tube. Bars = 50 µm. D,
Estimation of the quantity of RNA. RNA isolated from 20 (Lane1) and 2 (Lane 2)
captured whole ovule sections (5µm thick) was used for RT-PCR; Lane 3-5, Stage
4 ovary RNAs (25, 5 and 1 ng, respectively) were used as controls for quantity
estimation. E, Quantification of amplified RNA. RNA isolated from 2 whole ovule
sections was subjected to one round of RNA amplification followed by RT-PCR
using 2 µl of amplified product (Lane 1). Lane 2-4, Stage 4 ovary RNAs (5, 1 and
0.2 ng, respectively) were used as controls for quantity estimation. F, Assessment
of genomic DNA contamination in the amplified RNA samples. RT-PCR was
carried out using equal quantities of aRNA from sporophytic ovule (SO) cells,
aposporous initial (AI) and early aposporous embryo sac (EAEs) samples, and
genomic (g) DNA. G, Validation of aRNA quality derived from LCM-harvested cells
using Hieracium genes with known ovule cell-type expression profiles. Expression of
HDMC1 and HFIE in Hieracium ovules has been characterized by RT-PCR and in
situ hybridization (Okada et al., 2007; Rodrigues et al., 2008).
Supplemental Figure S2. Validation of RT-PCR data obtained for low ovary-expressed genes
in LCM samples by quantitative real-time PCR. The three AI cell detected genes (class II,
Figure 1C; ID 9.45 RD22, ID 24.04 NLR and ID 27.18 LOX, Table S2) and two genes from
class III (Figure 1C; ID 9.20 BAM1 and ID 9.10 CKX) and one from class IV (Figure 1C, ID
25.03 S-protein; Table S2) were chosen for quantitative PCR analysis. Gene expression levels
relative to the reference HUBQ gene (Rodrigues et al., 2008) are plotted on the y-axis for each
Supplemental Figure S7. Differentially expressed Arabidopsis homologs of H.praealtum
contigs from AI cell, SO cell and EAE sac data sets. Red dots indicate relative expression in
the FG2-4 vs ovule (late) and blue dots indicate the relative expression in the nucellus vs whole
ovule (early).
Cell type a Ovule b SO cells AI cells c EAE sacs c
Number of “cells” 500 2,500 270 100 Original RNA d 0.1-0.2 ng 0.5-1 ng 0.1-0.2 ng 0.1-0.2 ng 1st amplification e 600 ng 1.3 µg 646 ng 513 ng 2nd amplification f
N.A 189 µg 154 µg 126 µg
f, 10% (10 µl) of the first round amplified material was used for the second amplification. N.A, not available.
Supplemental Table S1 . Isolation of RNA from laser captured H. praealtum R35 apomictic ovule cell types and RNA yields after amplification.
a, Whole ovules, clusters of somatic ovule (SO) cells, aposporous initial (AI) cells and 2-4 nucleate early aposporous embryo (EAE) sacs were collected by LCM from 5 µm sections. b, Two LCM ovule sections estimated to contain 500 "cells" in cross section were used for RNA isolation and amplification.
c, This represents the number of individual AI “cells” or EAE sacs cut from 5µm thick sections. d, Quantity of original RNA was estimated using an indirect PCR method to amplify a control gene from a known amount of ovary RNA. e, Quantity of RNA after the first round of amplification was determined using a Nanodrop spectrophotometer.
Identifier Protein E-value SO AI EAE Forward Reverse
2.07 AT1G74770.1 unknown protein 2.00E-47 GO673075 VIATTACATGACCAC
ACGCAAG
AGGTAACGAGGAA
AGGATTC
2.11 AT3G56740.1
ubiquitin-associated
(UBA)/TS-N domain-
containing protein
2.00E-37 GO673077 + + + IACAACCCCTCCG
ATTAATTG
GTAAGCACACTCA
AAGTTGG
2.13 AT4G37190.1 unknown protein 4.00E-33 GO673078 VIICCAAAACCCCAAT
TTCGAAG
TTTATCAGGTTTC
GTGTCCG
2.18 No hits GO673081 IIIAACAACCAAGACG
TTCCAAC
TTTGAAGCCTCTA
TGCTGTC
2.20 AT5G59380.1MBD6 (methyl-CpG-binding
domain 6)2.00E-04 GO673083 I
GGGCAATTCTGTC
ATTTGTC
CTGGACAGAAATT
CAGATCC
2.21 AT3G04380.1SUVR4; histone-lysine N-
methyltransferase3.00E-09 GO673084 I
TAGGAAAGCTCTA
ATGCTGC
GATGGAACATGCT
TATCCTG
4.06 AT3G62770.1
AtATG18a (Arabidopsis
thaliana homolog of yeast
autophagy 18 (ATG18) a)
2.00E-08 GO673086 IATAGATTGCGATC
AATGGGG
GATGACTCAACTG
GAATATC
4.07 AT2G27880.1 argonaute protein (AGO5) 3.00E-05 GO673087 VITTCCATTGACCAA
TGGATGG
AAGAGCGTAGAAG
GTGAAAG
4.09 AT4G21710.1
EMB1989, RPB2, NRPB2 |
NRPB2 (EMBRYO
DEFECTIVE 1989); DNA
binding
1.00E-88 GO673089 IGTACGATACCGAA
TACGTTC
AAGAACGTCTCTT
TGACCAG
4.12 AT1G68720.1cytidine/deoxycytidylate
deaminase family protein2.00E-25 GO673090 I
TGTAATGGGAAGA
CACGAAG
AGAAGCTAAAAAG
GGTGCTG
4.18 No hits GO673092 ITCGCGTATTTAAG
TCGTCTG
TTTTGATCCTTCG
ATGTCGG
4.19 No hits GO673093 ICATAGCTTTTAGA
AGACGCG
GACAGAAATTGGG
ACAGAAG
4.20 AT5G19300.1 expressed protein 2.00E-17 GO673094 IIITGGCTAAAAAGAA
27.30 AT3G03940.1 protein kinase family protein 2.0E-19 GO673163 VGTACACTTTTCCC
TTACACC
GAAACACAGGAGA
CATTACG
27.34 AT3G23640.1
HGL1 (HETEROGLYCAN
GLUCOSIDASE 1);
hydrolase, hydrolyzing
O-glycosyl compounds
2.0E-48 GO673164 IIIGGTTTTTGAACCG
ATTCTGG
TTGTTGATAGCCC
AATGACC
27.35 AT1G68530.1
CER6, G2, POP1, CUT1 |
CUT1 (CUTICULAR 1);
acyltransferase
6.0E-38 GO673165 IACGAAGCAACAAC
GTGAAAG
GTTGAAGCATCCA
GAATGAC
27.37 AT1G03280.1
transcription initiation factor
IIE (TFIIE) alpha subunit
family protein / general
transcription factor TFIIE
family protein
2.0E-53 GO673166 IACATCTCTAGGAA
GAACACC
GTTCGGGAAATGT
GGATTTG
HUBQ AT4G05320.1 UBQ10NM_001084
884 I
ACTCCACTTGGTC
TTGCGTCT
AGTACGGCCGTC
TTCAAGC
HFIE (D36) AT3G20740.1
FIS3, FIE1, FIE | FIE
(FERTILIZATION-
INDEPENDENT
ENDOSPERM 1); nucleotide
binding / transcription factor
e-172 EU439051 ICCAGGAGAGGGC
ACAGTTGATA
GGGCTAGTTTGCA
ATTCCCATA
Table S2 (continued)
HRBR AT3G12280.1
RBR, RB, RBL1, RBR1 |
RBR1 (RETINOBLASTOMA-
RELATED 1)
0 EU439049 VICATGTGTTGGAGA
GAGCACACA
ACTTGATGAAGCG
GGACCTTTC
HDMC1 AT3G22880.1
DMC1, ATDMC1 | ATDMC1
(RECA-LIKE GENE); ATP
binding / DNA-dependent
ATPase/ damaged DNA
binding
e-163 EF530197 - - - -CAGCTGGCTCAC
ACTCTCTG
TCAAGTACAGCTC
CAGCATCC
a, The most similar Arabidopsis genes searched by blasts with TAIR database (http://www.arabidopsis.org/index.jsp) are listed with AGI identifier number, similarity and E-value.
b, Accession number of the Hieracium gene sequence deposited on GenBank database
c, RT-PCR anaylsis using aRNA from LCM samples. SO, Somatic ovule cells; AI, aposporous initial cells; EAE, embryo sac; -, no expression
d, Class defined by expression pattern (see text and Figure 1). -, no expression in any cell type
e, Primer sequences used for RT-PCR analysis
Pfam Domain ID Pfam Domain GO ID GO Description Ontologya Number in
input listb
Number in
BG/Refb p-value
c
SO vs AI
Enriched in SO
PF03095 PTPA Phosphotyrosyl phosphate activator (PTPA) protein GO:0019211 phosphatase activator activity F 5 0 3.2E-03
Supplemental Table S6. Discriminatory gene annotations associated with significant GO terms found through nested GO enrichment analysis between H. praealtum ovule cell types and comparison
with Arabidopsis ovule genes on Agilent 4x44k arrays. Blue dot (●) indicates detectable expression, gray identifies genes not present on the array, asterisk (*) highlights genes contributing to more than 1 GO