Gari Bay

Garibay 1

Osteogenesis Imperfecta: The Brittle Bone Syndrome

By Edgar Garibay

Osteogenesis Imperfecta (OI) is an autosomal dominant genetic disorder that affects the

formation of bone in approximately 1 in every 10,000, regardless of the ethnic group [1]. OI is

also commonly referred to as brittle bone disorder [1]. There are about four different types of OI

that are caused by a mutation in COL1A1 and COL1A2 genes that leads to the production of low

amounts of type I collagen [26]. The severity of the disease depends upon the site of the

mutation or the lack of type I collagen that is produced. Type I collagen is the main component

of tendon and bone and it is also vital for the tensile strength of bone [3].

Patients that suffer from OI are born with a lack of collagen, or the quality of collagen

constructed is poorer than normal. Since collagen is essential for bone formation, defective

collagen causes patients with this condition to have weak or fragile bones. As a result, OI

patients can suffer anywhere from a few to hundreds of bone fractures throughout their lifetime,

making this disease extremely variable in its phenotypic expression [1]. Of the different types of

OI, some patients express mild symptoms such as suffering from bone fractures, and can live a

normal lifestyle; however most of them are wheelchair bound at an early age. Other patients

who inherit the lethal form of OI die in their mother’s womb (perinetal) [1]. OI patients can also

suffer from weak muscles, abnormal tooth teeth, bluish coloring in the sclera of the eye, curved

spine, hearing loss, short stature, a triangular face, respiratory problems, and hearing loss [1 &

16]. Typically in an autosomal dominant disorder, only one copy of the abnormal gene is passed

from one of the parents to the offspring. This means that if one of the parents has the mutation

and has OI then there is a 50 % chance that any one of their offspring will inherit the mutated

Garibay 2

gene and express the genetic disease, such as the case of osteogenesis imperfecta. Most often, an

offspring that is affected by an autosomal dominant disorder is produced by the union of a

normal parent (aa) with an affected heterozygote (Aa) [1].

To understand the importance of osteogenesis imperfecta it is important to be familiar

with process and importance of collagen type I. In the cases referenced, I [1, 5, 8, 9, 10, 13, &

14] OI is caused by defect in type I collagen. Collagen is classified under the category of fibrous

proteins whose polypeptide chains are arranged in long strands or sheets [3]. Fibrous proteins

develop rod or wire like shapes that provide strength and or flexibility to structures where they

occur [3]. These proteins are made up of simple repeating amino acid residues that form a

secondary structure (left handed helix) [3]. Fibrous proteins are normally water-insoluble, which

means that the structure has a high concentration of hydrophobic amino acid residues both on the

inside of the protein and on its surface [3]. This implies that supramolecular structures (like

collagen) are made by packing similar hydrophobic polypeptides inside of the helix [3].

Collagen is a group of twenty six proteins that are the most abundant proteins in

vertebrates [2]. Collagen accounts for nearly 25 % of the body protein content and it is the major

structural protein of the extra cellular matrix [3]. The importance of collagen is that it provides

highly organized fibrous matrix in connective tissues that include: bone, cartilage, ligament,

tendon, dermis and dentin [8]. Each collagen type has its own function or set of functions. Of the

twenty-six types of collagen, the most studied collagen is type I [2]. The collagen helix is left

handed and has three amino acid residues per turn of the helix. Thus the basic subunits of

collagen in the alpha chain include trinucleotide repeating patterns of Gly--Xaa--Yaa, where Gly

is glycine, Xaa usually represents proline predominantly, but sometimes lysine is present and the

Yaa position corresponds to 4-hydroxyproline [2 & 3]. The mixture of glycines in every third

Garibay 3

amino acid repeat (makes the alpha helix highly flexible) and the existence of proline and 4-

hydroxyproline (provides tight turns on the alpha helix) allow the three alpha chains of collagen

to come together to form a right handed super helix [2 & 3]. Collagen type I makes a coiled-coil

secondary structure that consists of two alpha 1 chains and one alpha 2 chain that wraps around

one another like a strand on a rope forming a right handed super helix [2]. The genes that encode

the two alpha 1 chain and the one alpha 2 chain are each single genes located on chromosomes

17q21.3-q22 (COL1A1) and 7q21.3 (COL2A2) [2, 4 & 8]. Both of these alpha chains contain 52

exons that are dispersed throughout the COL1A1 or COL1A2 genes [8 & 13]. Of the 52 exons,

exons 7 to 48 encode the triple helix domain that contains 338 uninterrupted trinucleotide repeats

Gly--Xaa--Yaa repeats (1014 amino acids long) [8 & 13]. The COL1A1 and COL1A2 genes are

virtually identical; except for the exon domain 33 and 34 are fused in COL1A and therefore are

referred to as exon 33/34. On the other hand, the exon on domain 33 and 34 remain separated in

COL1A2 [8]. Another subtle difference between both of these genes is that COL1A1 is much

smaller 18 kilo base pairs than COL1A2 38 kilo base pairs since COL1A2 contains larger introns

[8] (See Figure 1).

Garibay 4

Figure 1: COL1A1 and COL1A2 Genes

Description:

A) COL1A1 gene is located on chromosome 17q21.3-q22.

B) COL1A2 gene is located on chromosome 7q21.3.

Image and description obtained Genetics Home Reference [22 & 23].

Garibay 5

Type-1 collagen chains are usually synthesized in what is called pre-pro-alpha-chains in

the endoplasmic reticulum [8]. In this process, 22 amino acid residues are removed to yield a

product of pro-alpha chains that contain globular N-terminal and C-terminal propetides [9 & 14].

The pro-alpha chains undergo many processes of association, registration and disulfide bonding

in the endoplasmic reticulum (ER) [9]. It is the two pro alpha 1 (I) and one pro algha 2 (I) chains

first join together via interaction between C-propeptides The association of these chains are

stabilized by the construction of interchain dulsulfide bonds that begin at the carboxy end of the

helix and progress to the amino terminal as the chains are being constructed in a zipper-like

fashion [1 & 13]. At the same time proline and lysine undergo hydroxylation through specific

hydroxylases to produce 4-hydroxyproline residues and hydroxylysine residues which are further

glycosylated by sugar transferases [1 & 13]. It is the hydroxyl groups in the hydroxyproline that

help connect the three alpha chains by forming hydrogen bonds [1].

The above mentioned processes guarantee the proper alignment of amino acid residues

that are necessary for the formation and propagation of the procollagen triple helix [8]. The pro-

alpha chains also undergoes co-and posttranslational modifications that involve nine enzymes

that reside in the ER. Other enzymes that are involved in this process are molecular chaperones

HSP47, BiP, GRP94 and PDI (protein disulphide isomerase) which are believed to assist in the

proper folding of pro-collagen [9]. Molecular chaperones are usually synthesized in response to

increased temperatures or other stresses arising in the cell. Thus it is believed that not only do

molecular chaperones aid in the folding of newly synthesized proteins, but some can be involved

in repairing potential damage caused by improper folding of a particular protein [10]. What has

been discovered from mutational studies about the proper folding and assembly of procollagen is

Garibay 6

that they interact longer with molecular chaperone proteins in the E. R.; these chains eventually

will degrade because of the continual contact with molecular chaperones [9]. Once these

processes have occurred, specific proteinases (N-proteinase and C-proteinase) cleave the carboxy

and amino terminal propepeptides to produce a mature type I collagen (three alpha helices) that

is sent to the Golgi apparatus and is secreted out into the cytoplasm [1, 8, 11 & 12]. Outside of

the cell the type I collagen then self-aggregates into fibrils (highly organized) that are stabilized

by intermolecular cross links. The aforementioned cross-links are produced from oxidative

deamination of lysine and hydroxylysine residues [13]. These highly ordered fibrils provide

mechanical strength for connective tissue (See figure 2 & 3).

Garibay 7

Garibay 8

Garibay 9

Figure 3: Procollagen to Collagen Triple Helix Formation

Description:

A) Represents the intracellular and extracelluar steps involved in the synthesis, processing and

assembly of type I collagen molecules.

B) Represents a mutation in one of the pro-alpha chains. Subsequently this produces a kink in in

the defective fibril formation. The (X) represents a mutation.

Image and description obtained Gajko- Galicka, A. [13].

Garibay 10

Mutations that lead to defects in type I collagen (resulting in OI) can be set up into two

broad categories of genetic defects: One of these categories involves dominant negative mutation

in the COL1A1 or COL2A1 [8]. The second category involves null mutations that affect

COL1A1 [8]. The abnormalities that are produced by dominant negative mutations involve the

sequence of the carboxy propeptide or the procollagen alpha chains. A dominant negative effect

produces a gene product that is not only non-functional (procollagen type I alpha chain(s)), but

the effect also inhibits the normal allele from producing the functional protein product [1 & 15].

As a result the abnormal type I collagen is synthesized as seen in figure 4c. Typically dominant

negative mutations are more detrimental than null mutations [15]. The most typical dominant

negative mutation is the single amino acid substitution of glycine within the triple helix [5, 8 &

13]. Substitution of glycine in the triple helix can produce a wide range of severities for OI. For

example, some mutations in glycine substitution can produce 50 % of the abnormal type I

collagen, while other mutations in glycine substitution can produce 75 % of the abnormal

product, as seen in figure 4c [13].

There are several important factors that determine the severity of OI in an individual: the

type of substitution that occurs, the position in the chain, the sequences surrounding the mutation

and the chain in which the substitution occurs [8]. The DNA triplet codon that makes up the

glycine residue in the procollagen alpha chain usually consists of sequence GGN [8]. Single

amino acid substitution from glycine in the procollagen chain usually occurs in the first two

nucleotides of the triplex codon (GGN) with substitutions of the following eight amino acids:

alanine, cysteine, argentine, aspartic acid, cysteine, glutamic acid, serine, trytophan or valine [5

& 8]. Substitutions of cysteine are the most abundant, while substitutions of alanine are rarely

Garibay 11

seen. There is strong evidence that suggests that a majority of the mutations occur in the first

base pair of the triplet codon, instead of the second one [8 & 13]. Furthermore, the substitution

of a single amino acid can occur in any of the three alpha chains. As a result the severity of OI

can range from mild to lethal in the phenotypic expression of an individual as seen in figure 4c

[5, 8 & 13]. For example, type I OI is the mild form that results from a null mutation. A null

mutation occurs when the defective gene is not transcribed into RNA and/or is not translated into

a functional protein product [25]. Therefore, patients with type I OI usually have the normal type

I collagen, but only produce half of the amount of collagen as seen in figure 4b [13]. On the

other hand, a patient with Type II, III or IV OI (moderate to lethal types of OI) has low levels of

the abnormal type I collagen that is caused by dominant mutations as previously discussed [6].

Another subtle difference is that lethal OI mutations are usually found in the C-terminal half of

the procollagen alpha chains, while non-lethal mutations occur around the N-terminus [19]. In

the work conducted by Yang, et. al, 1997, they explored the mutation site of non-lethal and lethal

OI. They designed peptides in a specific region on one of the alpha chains (I) that included a

non-lethal site substitute of serine to glycine at residue 901 and lethal substitution site of serine

to glycine at residue 913 [19]. The normal sequence of the designed peptide included residues

892-909 which contained four Gly--Xaa--Yaa triplets formed a stable triple helix. The first

substitution was introduced in glycine 901 with a serine residue. This substitution produced only

50 % loss of the triple helix formation which had a small effect on the folding of the triple helix

[19]. A second peptide of normal sequence was designed with the residues 904-921. Similarly to

the results of the first designed peptide, the second one formed a stable triple helix. However,

unlike the previous peptide, the substitution of serine for glycine resulted in a much greater

decrease of triple helix formation, which had a greater impact on the folding of the triple helix.

Garibay 12

This study illustrates the severity of glycine substitutions that result in the different forms of OI

(lethal vs. non-lethal).

The mutations that are present in many OI patients are usually exclusive to the affected

individual or family members [8 &13]. In the referenced experiments [8 &13] several

substitutions of glycine have been determined. For example, a substitution in glycine 1009,

glycine is substituted by serine or valine [8]. Substitutions in glycine in 832 consist of aspartic

acid or serine and glycine 415 is substituted by cysteine or serine [8]. A single amino acid

substitution study of two unrelated children with severe OI and two other unrelated children with

lethal perinatal OI found that dominant mutation of glycine are not exclusive to each individual

or family members as [8] had previously suggested [14].

The formation of the triple helix starts at the carboxy terminus end and continues its

normal rate to the amino terminus; the formation of the triple helix can only be interrupted

(stopped) when it encounters a substitution of the normal amino acid sequence [8]. Consequently

further elongation of the helix is postponed in the amino segments of the chain that remain in its

nascent state for long periods of time [8].

The second category of genetic defects involves null-mutations that occur predominantly

in the COL1A1 gene, where the mutant allele cannot produce pro-alpha chain or if it does

produce one it is usually functionless [8 & 13]. One null mutation in COL1A1 allele results in

half of the mature type I procollagen synthesized. This type of defect usually results in mild OI

(See Figure 4). The exact molecular mechanism of a null mutation is not exactly known, but

there are a few studies that propose a possible mechanism [8, 20 & 21]. This mechanism

suggests that the amount of mRNA for the COL1A1 gene is not sufficient for translation of

procollagen I. [21]. In the study conducted by Willing, M. C., et. al., 1996, they contend that

Garibay 13

most common cause of a translational defect in procollagen I has to do with pre-mature

termination in which RNA is altered [21].

To truly understand how RNA alteration occurs, Hartikkam H., et. al. 2004 studied null

mutations of exons 11-49 of the COL1A1 gene. Samples for PCR analysis were obtained from

dermal fibroblast or lymphocytes of affected individuals (type I OI), unaffected family members

and unrelated controls [21]. The PCR product was then electrophoresed in 0.8% NuSieve

solution. Individual bands from the gel were cut and melted. The samples were then submitted

for DNA sequencing using Sanger method of dideoxy chain termination method. The PCR

product (exon 11-49) was analyzed by Single-Strand Conformation Polymorphism (SSCP) to

search for possible null mutations [21]. SCCP is a procedure that separates single-stranded

nucleic-acids through based on slight differences in the sequence [24]. This result in different

mobility of the PCR samples in the gel [24].

In a separate procedure, dermal fibroblast samples taken from OI patients were used

to isolate nuclear and cellular RNA [21]. To complete, this process the dermal fibroblasts were

incubated with ascorbic acid for 4 days (medium was changed daily). Isolation of cellular RNA

was accomplished by guanidinium thiocyanate-phenol-chloroform method [21]. Isolation of

nuclear RNA on the other hand, was accomplished by washing the fibroblast samples with PBS

and subsequently adding the sample to a lysis buffer solution [21]. The experimental results

concluded that mutation in the RNA (nuclear and cellular) reduced the amount COL1A 1 mRNA

that is available to produce type I collagen which results in the mild phenotypes of OI type I [21]

(see figure 4 for possible mutations and collagen product).

OI can be diagnosed in the womb as early as 14-18 weeks through the use of ultrasound

that may identify bone lesions or abnormalities [6]. Other prenatal testing methods include

Garibay 14

chorionic villus sampling and amniocentesis that also be used to determine genetic mutations by

obtaining a cell sample. Often times it is not always possible to diagnose OI before birth. In

children, OI usually isn’t noticed until the child begins to walk or if severe fractures are noticed

in the child. Before proper diagnosing methods were fully developed, it was once thought that

children with a number of fractures were abused, instead of being diagnosed with OI [17].

Although some parents have been wrongly accused of child abuse, it is important that the doctor

makes a thorough diagnosis by looking for changes in the bone fractures (i.e. X-rays) that may

develop in the child to determine if it is OI or child abuse [17]. For cases that go undiagnosed in

early childhood, doctors look at other prominent clinical features that include: pain in various

parts of the body, (i. e. spine, hip, feet) difficulty walking, and fractures that typically occur in

the long bones of the body. It is often difficult to diagnose IO exclusively on clinical features [6].

Furthermore, other diagnostic tests such as collagen analysis or DNA sequencing are needed to

help confirm a diagnosis for OI [6]. Currently there is no cure for OI, but the goal for these

patients is to provide treatment that can prevent deformities or fractures of the bone [7]. Some of

the treatments can include: dental procedures, physical therapy, surgery, taking care of the

fractures, and rodding (inserting a metal plate or bar in the long bones to stabilize it and prevent

it from deforming).

Garibay 15

Figure 4: Molecular Mechanism of Osteogenesis Imperfecta

Description:

A) Normal assembly of type I collagen.

B) Mutation with a decrease in the production of collagen type I (50 %) is due to a null

mutation, resulting in osteogenesis imperfecta.

C) Deforming osteogenesis imperfecta depends on the gene in which the mutation occurs.

The abnormal molecule is reflected in the phenotype.

The (X) represents mutations in the procollagen chains; the vertical lines symbolize post-

translational modifications.

Image and description obtained Gajko- Galicka, A. [13].

Garibay 16

References

1. Jorde, L. B., Carey, J. C., Bamshad, M.J. & White, R. L Medical Genetics. Mosby

3rd edition. 2006.

2. Dalgleish, R. The Human Type I Collagen Mutation Database. Nucleic Acids Research, 25,

181–187. 1997. 26 Feb. 2008

3. Cox, M. M. & Nelson, D. L. The Principles of Biochemistry. Lehninger. 4th edition. 2005.

4. HUGO Gene Nomenclature Committee. <http://www.genenames.org/data/

hgnc_data.php?match=COL1A2.> July 16, 2007. 26 Feb. 2008

5. Yang, W., Battineni M. L., & Brodsky B. Amino Acid Sequence Environment Modulates the

Disruption by Osteogenesis Imperfecta Glycine Substitutions in Collagen-like Peptides.

Biochemistry, 36, 6930-6935. 1997. 26 Feb. 2008.

6. Osteogenesis Imperfecta Foundation. Fast Facts on Osteogenesis Imperfecta. .2007.

26 Feb. 2008. <http://www.oif.org/site/PageServer?pagename=FastFacts>.

7. University of Maryland School of Medicine. Osteogenesis Imperfecta.

<http://www.umm.edu/bone/oi.htm> 2004. 30 Apr 2008.

8. Cohn, W. E. & Moldave, K. Progress in Nucleic Acid Research and Molecular Biology.

Academic Press Inc. 47, 1-397, 1994.

9. Lamande S. R., & Bateman, J. F. Procollagen Folding and Assembly: The Role of

Endoplasmic Reticulum Enzymes and Molecular Chaperones. Cell and Developmental

Biology, 10, 455-464, 1999.

10. Evstigneeva Z. G., Solov'eva, N. A. & Sidel'nikova, L. I. Structures and Functions of

Chaperones and Chaperonins (Review). Applied Biochemistry and Microbiology

Garibay 17

37, 1-13, 2001. 3 March, 2008.

11. Lammi, L. Type I and III Procollagen Propeptides in Sarcoidosis, Fibrosing Alveolitis and

Asbestos-Related Lung Diseases. American Journal of Respiratory Critical Care

Medicine 159, 818-823, 1999. 3 March, 2008

12. Dombrowski K. E., Voge1 B. E. & Prockop D. J. Mutations That Alter the Primary

Structure of Type I Procollagen Have Long-Range Effects on Its Cleavage by

Procollagen N-Proteinase. Biochemistry, 28, 7 107-7 112, 1989. 31 March, 2008

13. Gajko- Galicka, A. Mutations in Type I Collagen Genes Resulting in Osteogenesis

Imperfecta in Humans. Acta Biochimica Polonica, 49, 433–441, 2002. 31 March, 2008

14. Pruchno, C. J., Cohn, D.H., Wallis, G.A., Willing, M. C., Starman, B.J, Zhang X. M. &

Byers, P.H. Osteogenesis Imperfecta Due to Recurrent Point Mutations at CpG

Dinucleotides in the COL1A1 Gene of Type I Collagen. Human Genetics, 87, 33-40.

1991.

15. Genetics Home Reference. Dominant Negative Mutation. March 21, 2008. 31 March

2008. < http://ghr.nlm.nih.gov/glosscry>

16. American Academy of Orthopedic Surgeons. Osteogenesis Imperfecta. August, 2007

31 March 2008. <http://ghr.nlm.nih.gov/glossary=dominantnegativemutation>

17. Chevrel, G. & Le Merrer M. Osteogenesis Imperfecta. Orphanet Encylopedia. 2004

31 March 2008. <http://www.orpha.net/data/patho/GB/uk-OI.pdf>

18. Wheeless, C. R. Type I Collagen. Textbook of Orthopedics. 31 March 2008

<http://www.wheelessonline.com/ortho/type_i_collagen>

19. Yang, W., Madhavi, L. B. & Brodsky, B. Amino Acid Sequence Environment Modulates

the Disruption by Osteogenesis Imperfecta Glycine Substitutions in Collagen-like

Garibay 18

Peptides. Biochemistry, 36, 6930-6935. 1997.

20. Hartikka, H., Kaija, K., Korkko, J., Kaitila, I., Gre´nman, R., Pynnonen, S., Hyland, J. C.,

& Ala-Kokko, L. Lack of Correlation between the Type of COL1A1 or COL1A2

Mutation and Hearing Loss in Osteogenesis Imperfecta Patients. Human Mutation, 24,

147-154. 2004.21. Willing, M. C., Sachi, D., Slayton, R. L., & Roberts, E. J. Premature Chain Termination Is

a Unifying Mechanism for COL1Al Null Alleles in Osteogenesis Imperfecta Type I Cell

Strains. American Journal of Human Genetics, 59, 799-809. 1996

22. Genetics Home Reference. COL1A1. 14 Apr 2008. 16 April 2008.

<http://ghr.nlm.nih.gov/gene=col1a1>.

23. Genetics Home Reference. COL1A2. 14 Apr 2008. 16 April 2008.

<http://ghr.nlm.nih.gov/gene=col1a2 >.

24. Davidson University. Single-Strand Conformation Polymorphism (SSCP). 16 April 2008.

<http://www.bio.davidson.edu/Courses/Molbio/MolStudents/spring2003

Parker/method.html>.

25. Null Mutation. Webster’s New World Medical Dictionary. 1 Jan 2001. 16 April 2008.

< http://medterms.com/script/main/art.asp?articlekey=4599>.

26. Genetics Home Reference. Osteogenesis Imperfecta. 25 Apr 2008. 30 Apr 2008

<http://ghr.nlm.nih.gov/condition=osteogenesisimperfecta>.