Temporal variation in the deposition of different types of collagenwithin a porous biomaterial implant

Jacinta F. White,1 Jerome A. Werkmeister,1 Teresa Bisucci,2 Ian A. Darby,2 John A. M. Ramshaw1

1CSIRO Materials Science and Engineering, Bayview Avenue, Clayton, Victoria 3169, Australia2School of Medical Sciences, RMIT University, Bundoora, Victoria 3083, Australia

Received 15 October 2013; accepted 31 October 2013

Published online 16 November 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.a.35027

Abstract: The deposition of new collagen in association with

a medical implant has been studied using expanded polytet-

rafluoroethylene vascular replacement samples implanted

subcutaneously in sheep, for up to 28 days. New type I colla-

gen mRNA synthesis was followed by in situ hybridization,

while the accumulation of new collagen types III, V, VI, XII,

and XIV was followed by immunohistochemistry. All the col-

lagen detected in the pores of the implant were newly depos-

ited at various times after implantation and were not due to

any pre-existing dermal collagen that may have been present

around the implant. Collagen deposition was seen initially

surrounding the implant and, with time, was seen to infiltrate

within its pores. In situ hybridization showed that the major-

ity of infiltrating cells had switched on mRNA that coded for

type I collagen production. Histology showed that cellular

infiltration increased with time, accompanied by increasing

collagen deposition. The deposition of different collagen

types happened at different rates. The type V and VI colla-

gens preceded the major interstitial collagens in the newly

deposited tissue, although at longer time points, detection of

type V collagen appeared to decrease. After disruption of the

interstitial collagens with enzyme, the “masked” type V colla-

gen was clearly still visible by immunohistochemistry. Little

type XII collagen could be seen within the porous mesh,

although it was seen in the surrounding tissues. By contrast,

type XIV was seen throughout the porous structure of the

implanted mesh, with less being visible outside the material

where type XII was more abundant. VC 2013 Wiley Periodicals,

Inc. J Biomed Mater Res Part A: 102A: 3550–3555, 2014.

Key Words: collagen, immunohistochemistry, in situ hybrid-

ization, porous material, cell, material interaction

How to cite this article: White JF, Werkmeister JA, Bisucci T, Darby IA, Ramshaw JAM. 2014. Temporal variation in the deposi-tion of different types of collagen within a porous biomaterial implant. J Biomed Mater Res Part A 2014:102A:3550–3555.

INTRODUCTION

Cell infiltration and collagen deposition are important proc-esses in the host tissue integration and efficacy of porousbiomaterials, including those used as supports in tissueengineering. While the characteristics of cellular infiltrationhave been widely studied, much less is known about theprocesses and timing involved in collagen synthesis, deposi-tion and fibrillogenesis, including the order in which varioustypes of collagen assemble in newly formed tissues.

Overall, some 28 distinct collagen types have been iden-tified in human tissues, with some of these also having vari-ous additional isoforms or splice variants.1 All share thepresence of the characteristic triple helical motif, whichcomprises a (Gly-Xaa-Yaa)n repeating sequence, in differentproportions in their structures.1,2 Only a few collagens,notably types I, II, and III collagens, are found in largeamounts, where they comprise the principle fibrillar compo-nents of tissues. All other collagens are normally found inonly small amounts, although they may be at high concen-tration in their particular tissue niche.

Originally, an expanded polytetrafluoroethylene (ePTFE)tube implantation had been introduced as a miniature

model for evaluation of wound healing potential, especiallyin humans.3 It allowed ingrowth of connective tissue intothe porous structure to be examined by histology and byquantitative measurement of total collagen synthesis,assessed by hydroxyproline content.3 Previously, we used avariant of this model, which involved the subcutaneousimplantation of an ePTFE tube, to study the development ofnewly deposited, individual banded collagen fibrils that con-tained types I and III collagens that were formed within thelumen of the tube.4

In this study, we have used in situ hybridization (ISH)and immunohistochemistry in combination with this woundhealing model to examine the individual collagens and thetime sequence of their deposition within the porous wall ofthe implanted ePTFE tube. Previous studies have used thismodel system to examine the effects of physiological andother parameters on the wound healing index monitored bythe total collagen accumulation within the porous struc-ture.3,5,6 This, more detailed study illustrates the additionalinformation that can be obtained from this model. Under-standing the nature and type of collagen deposition withinand around a medical implant is important in the

Correspondence to: J. A. M. Ramshaw; e-mail: John.Ramshaw@csiro.au

3550 VC 2013 WILEY PERIODICALS, INC.

understanding of performance. It is also important in thedevelopment of tissue engineered constructs, where it ishighly preferable that the new tissue closely mirrors of thenative tissue that it is intended to replace.7

MATERIALS AND METHODS

ImplantsePTFE tubing (GoretexTM, W.L. Gore and Associates, Flag-staff, AZ) segments (4 mm ID 3 20 mm) were implantedsubcutaneously in the back of sheep. These samples wereremoved for examination after 6, 14, and 28 days. Collagenformation in the wall of the tube was examined by histology,immunohistochemistry and ISH.

HistologySections of the tube and adjacent tissue were fixed at eachtime point in 10% neutral buffered formalin overnight,dehydrated in ascending grades of ethanol, cleared in Histo-choice (Sigma, St Louis, MO) and processed through to par-affin wax by standard protocols. Sections 4-mm thick werecut, had the paraffin removed, and were stained using Har-ris’ Haematoxylin and Eosin.

In situ hybridizationISH for mRNA encoding collagen type I was performed usingUTP-33P detection using 4 mm paraffin sections collectedonto 3-aminopropyltriethoxy-silane treated slides, followingthe method of Darby and colleagues.8 The type I collagenriboprobe was 600 bp in length, derived from the rat type I(aI) collagen sequence, and was sub-cloned into pGEM3z.9

ImmunohistochemistrySections, 4-mm thick, were cut from frozen samples from eachtime point and were examined using murine monoclonal anti-bodies (MAb) to type I (5D8-G9/Col1),10 type III (2G8-B1/Col3),11 type V (1E2-E4/Col5),12 type VI (1E8-B7/Col6),13 andtype XII (378D5) (Enzo Life Sciences, Farmingdale, NY) colla-gens, and an affinity purified rabbit polyclonal antibody againsttype XIV collagen (SAB4503061) (Sigma, St Louis, MO). Sectionswere pre-treated with enzyme in some cases and then incu-

bated with a primary antibody for 16 h at 4�C. For anti-type Icollagen, pre-treatment was with 1 mg/mL pepsin (Worthing-ton) in 0.01M acetic acid for 2 min at 37�C, then washed threetimes with phosphate buffered saline, pH 7.4 (PBS), prior toincubation with purified 5D8-G9/Col1 at 0.073 mg/mL (1 in 30of 2.2 mg/mL stock). For anti-type III collagen, no pre-treatment was required and incubation was with 2G8-B1/Col3at 0.018 mg/mL (1 in 30 of 0.53 mg/mL stock). For anti-type Vcollagen, sections were examined with no pre-treatment orafter pre-treatment with 0.07 mg/mL bromelain (Sigma) in PBSfor 4 min at 37�C and then washed with PBS before incubationwith purified 1E2-E4/Col5 at 0.11 mg/mL (1 in 30 of 3.2 mg/mL stock). For anti-type VI collagen, the neat cell supernatantfor 1E8-B7/Col6 was used and no pre-treatment was required.For anti-type XII collagen, pre-treatment of ice cold methanolfixed sections was with 0.005% trypsin for 5 min at 37�C,according to the supplier’s recommendations, and the MAb(378D5; provided as ascites at 5 mg/mL) then used at a 1 in500 dilution. For anti-type XIV collagen, the polyclonal antibody(SAB4503061) was used at 1 in 50 of 1 mg/mL supplied. Sec-tions were then washed three times with PBS. For mouse pri-mary MAbs, interactions were visualized with Alexa-Fluor 488labeled goat anti-mouse IgG (H1L) (Life Technologies) diluted1 in 500 with PBS for 30 min at room temperature. For therabbit polyclonal antibody, the secondary antibody wasAlexa-Fluor 488 labeled goat anti-rabbit IgG (H1L) (Life Tech-nologies) diluted 1 in 500 with PBS for 30 min at roomtemperature. Sections were then washed three times in PBSand were mounted in Dako Fluorescent Mounting Medium(Dako Agilent, Australia). Sections were examined using anOlympus BX61 microscope system. Controls, including appro-priate isotype controls, were also examined.

RESULTS

HistologyHaematoxylin and Eosin staining showed that after 4 days acapsule had formed around the outside of the implant. Thiscapsule became better defined at longer time points. At 6days, there was evidence of cellular infiltration and new tis-sue formation into the porous implant [Fig. 1(A)], and the

FIGURE 1. Haematoxylin and eosin staining of ePTFE explant samples after (A) 6 days, and (B) 14 days. [Color figure can be viewed in the online

issue, which is available at wileyonlinelibrary.com.]

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | OCT 2014 VOL 102A, ISSUE 10 3551

extent and depth of this infiltration increased with time[Fig. 1(B)].

In situ hybridizationThe samples examined by H&E histology were also exam-ined by ISH, using a riboprobe toward type I collagen.8 Inthe 4-day explant, ISH showed active mRNA synthesis forcollagen around the outside of the implant in the region ofthe capsule formation, but none within the porous wall ofthe implant. At 6 days, ISH showed a limited number ofcells had infiltrated into the porous wall of the ePTFE andwere synthesizing mRNA for type I collagen, in addition tocontinued detection of mRNA in the surrounding capsule[Fig. 2(A)]. At later time points, including 14 [Fig. 2(B)] and28 days, there was an increasing amount of cellular infiltra-tion within the porous material, and increasing depths of

the implant, with associated type I collagen mRNA biosyn-thesis. ISH showed that most, if not all, cells were synthesiz-ing type I collagen mRNA at all time points, although ISHdoes not indicate the amount of type I collagen protein thatwas accumulating in the extracellular matrix.

ImmunohistochemistryThe nature of the collagen matrix that was deposited withinthe pores of the ePTFE implant was examined by immuno-histochemistry. These data showed that the deposition of dif-ferent collagen types took place at different times, ratherthan all at the same time. Type V and type VI appeared to bedeposited before the major interstitial collagens [Fig. 3]. TypeVI collagen was observed in the capsule that formed aroundthe implant at 14 days, along with infiltration mostly on theouter half of the implant [Fig. 3(A)]. By 28 days, the type VI

FIGURE 2. ISH staining of ePTFE explant samples for type I collagen synthesis after (A) 6 days, and (B) 14 days.

FIGURE 3. Immunohistochemistry examination of ePTFE explant samples after (A, B, C) 14 days and (D, E, F) 28 days for (A, D) type VI collagen,

(B, E) type V collagen, and (C, F) type III collagen.

3552 WHITE ET AL. DEPOSITION OF NEW COLLAGEN WITH A MEDICAL IMPLANT

collagen had infiltrated uniformly through the full thicknesson the implant, while its presence in the surrounding capsulewas still clear [Fig. 3(D)]. Type V collagen, in sections with-out enzyme pre-treatment, was most visible at the earliertime points. However, even at 4 and 6 days, there was littlestaining observed in the capsule around the implant. At 14days [Fig. 3(B)], the majority of staining was at deeper loca-tions within the porous structure of the implant, with lessstaining observed in the outer zone where infiltration wouldfirst occur. By 28 days [Fig. 3(E)], the staining for type V col-lagen had almost completely disappeared, with only a smallamount of staining observed at the deepest points. Type IIIcollagen showed strong staining, initially of the capsule, butthen throughout the implant. At 14 days [Fig. 3(C)] type IIIcollagen was clearly visible within the implant, with themajority in the outermost half. By 28 days [Fig. 3(F)] thestaining of type III collagen was strong throughout the fullthickness of the implant. Similarly, type I collagen (data notshown) showed a similar pattern of infiltration, consistentwith the ISH data, and also appeared to slightly lag behindthe infiltration of the type V and VI collagens.

The apparent disappearance of the type V collagen by28 days after the implant could be due to turnover or dueto interstitial collagen deposition that appears to follow thetype V collagen infiltration and which could mask the epi-tope recognized by the MAb to type V collagen. Sectionsfrom 28 day samples were treated with various enzymes tosee if the original type V collagen was still present andcould be unmasked. Treatment with bromelain gave thebest results [Fig. 4], and showed that the type V collagenwas still present and could be unmasked from the coverageby interstitial collagens.

Type XII and type XIV collagens showed different and dis-tinct responses in the present wound healing model of cellu-lar infiltration and new tissue formation into a porousbiomedical material. At 28 days, type XII collagen was readilyseen in the capsule that had formed around the implant, butwas barely visible as part of the new tissue that had formedwithin the porous implant [Fig. 5(A)]. To assess if theabsence of staining was due to interstitial collagen masking,sections were treated with trypsin but no type XII collagen

was seen within the pores, while the collagen in the capsulewas still readily observed, and had not been removed by thetreatment [Fig. 5(A)]. By comparison, after 28 days, type XIVcollagen was readily observed, and was present throughoutmost of the thickness of the pore region [Fig. 5(B)]. However,unlike type XII collagen, it was not readily observed in thenewly formed capsule region [Fig. 5(B)].

DISCUSSION

The tissue response to biomedical implants has previouslybeen studied using MAb. For example, studies on a range ofpolymer and metal implants showed that type III collagenwas present in all capsules,14 whereas the staining for typeI collagen was surprisingly absent for some polymer materi-als but present for titanium. Later studies on the depositionof type I and III collagens around small stainless steel andtitanium implants,15 showed the presence of both collagensin the newly formed capsules and suggested that although athicker capsule formed around the titanium, there were nosignificant collagen composition differences between thecapsules. More detailed studies using antibodies to types Vand VI collagens as well as to type I and type III collagenshave been reported for a porous, collagen based implant,the Omniflow Vascular ProsthesisTM 11,13,16 and haveincluded studies on both short-term17 and longer term (4year)18 new tissue formation. These studies showed theinfiltration of new tissue into the device and its

FIGURE 4. Immunohistochemistry examination of ePTFE explant mate-

rial after 28 days for type V collagen after treatment with bromelain.

FIGURE 5. Immunohistochemistry examination of ePTFE explant sam-

ples after 28 days for (A) type XII collagen, including treatment with

trypsin, and (B) type XIV collagen.

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | OCT 2014 VOL 102A, ISSUE 10 3553

accumulation over 4 years. They also suggested that types Vand VI collagens may be early markers for new tissue for-mation. Overall, these studies with porous and non-porousmaterials clearly showed the value of examining the individ-ual collagen type rather than only total collagen accumula-tion by histological or biochemical methods.

In this study, we have examined the rate and extent ofextracellular matrix tissue infiltration into a porous implant,visualizing the distribution of six different collagen typesincluding collagen types I, III, V, VI, XII, and XIV. We havealso used a better defined porous material, ePTFE, whichhas been used in previous studies as a well defined woundhealing model.3,5 This model has been used to track theeffects of anaemia, underlying clinical disease, angiogenicrestriction and psychological intervention on the total colla-gen content in wound healing.6 The model has also beenused to evaluate and define the type of cells involved in thetissue that infiltrates into the porous matrix using FACSanalysis of extracted cell populations.19 Previously, we haveused this model to extract collagen fibrils of defined maxi-mum age from the inner luminal volume of the implant, andshown that mature banded collagen fibrils had formedwithin 14 days that contained both the collagen I and colla-gen III within the same fibril.4

The main, interstitial, type I and type III collagens,which predominate in wound healing responses, were read-ily observed over the time period up to 28 days, with theexpression of type I collagen shown at early time pointsthrough use of ISH. Of more interest, potentially, was thebehavior of the less abundant collagen types. The infiltra-tion of type VI collagen was readily observed, and, as previ-ously noted,13 preceded the accumulations of the interstitialcollagens. This was also the case for type V collagen, whichis regarded as a regulatory fibril-forming collagen,20 throughthe early time points up to 14 days. However, beyond thispoint, type V was not as readily observed. However, treat-ment of the newly formed tissue with proteolytic enzymeled to the unmasking of this collagen revealing that it wasstill present in the tissue sample. This behavior is consistentwith type V collagen’s key role in the initiation of fibrilassembly,21 including as shown through studies on collagenV null mice.22 In normal fibril assembly in vivo, it has beensuggested that it may have a role as an initial thin core fila-ment that serves as a nucleation site for subsequent fibrilgrowth23 or that is more broadly distributed within a fibriland readily covered by other major interstitial collagensduring a role in fibril diameter control.24 Both models couldaccount for the masking of the type V collagen during fibrilgrowth, while allowing for it to be unmasked by mild prote-olysis. In relation to the performance of an implanted mate-rial, detection of type V collagen without unmasking in theearly wound healing phase and then later by unmasking,signifies that appropriate collagen synthesis and remodelinghas occurred.25

The other collagens examined, type XII and XIV collagensare members of the fibril associated collagens that haveinterrupted triple helices (FACITs) that are localizednear the surface of banded collagen fibrils.26 Although these

collagens have very similar structures, they play distinctroles in tissue formation.27 This is clear from the quite dif-ferent distributions observed in the present study, wheretype XII collagen was absent within the porous mesh, butwas present in the newly formed surrounding capsular tis-sues, while type XIV collagen was seen throughout theimplanted mesh, with little being visible outside the mate-rial where type XII was more abundant. Previously, studieshad mainly looked at the roles of these collagens duringdevelopment.28,29 For example, it was found that in bovinefoetal skin, type XIV collagen was not present at 9 weeks,but was significant from 19 weeks, such that its expressionwas distinct from the fibrillar collagens,29 yet prevalentwithin relatively dense dermal tissues.30 In human skinafter birth, type XII and type XIV collagen show changes inexpression patterns, with type XII collagen becoming preva-lent in structures such as hair follicles, where finer fibrilsmay be present, whereas type XIV was in the region of thelarger banded fibrils, which seem a prerequisite for theaccumulation of this collagen in dermis.31 In this context,type XIV collagen interacts with the fibril surface and regu-lates fibrillogenesis32 and hence may be involved in modu-lating the biomechanical properties of tissues.33 Thus, itsregulatory role in fibrillogenesis occurs later and seems dis-tinct from the relatively early the regulatory role of type Vcollagen. In the present studies, the association of type XIVcollagen is in accord with the development of a significantinterstitial collagen deposition within the porous scaffold. Itis not clear with the new tissue formation in and aroundthe implant why type XII collagen becomes exclusivelylocated in the newly formed capsule environment. However,it may reflect that in some newly formed capsules, the colla-gen tissue proximal to the implant consists of poorlyordered collagen with finer fibrils than found in dermal orsimilar tissue.34 Apart from dermis, the other tissues wheretype XII and type XIV collagens have been extensively stud-ied is the cornea and the lens capsule, where various distri-butions have been observed. For example, in the rabbitcornea, normal endothelial cells but not stromal cells havemRNA for collagen type XII.35 Collagen type XII has beenshown in corneal wound repair to accumulate to the woundbed, decreasing between 14 and 21 days.36 The temporaldistribution of collagen type XII and XIV up to 4 weeks hasbeen explored during capsular damage in a rat model. TypeXII and XIV collagens were absent in the uninjured eyes, butboth accumulated and co-localized beneath the lens capsuleby 4 weeks in the capsular model.37 In contrast, in ourstudy, at 4 weeks distinct distributions of type XII and typeXIV collagens were observed.

The model system used in the present study provides aneasy system to study new collagen deposition into a porousmedical implant, and allows the changes in collagen deposi-tion and availability at different time points after implant tobe readily explored. These data show that each of the sixcollagens examined in the present study follows a differenttimetable for deposition in the newly formed tissue. Thesystem has previously been used to examine the presenceof various cell receptors38 and could also prove useful for

3554 WHITE ET AL. DEPOSITION OF NEW COLLAGEN WITH A MEDICAL IMPLANT

studying other molecules, such as decorin or biglycan, thatare also involved in the regulation of deposition of newtissue.39

ACKNOWLEDGMENTS

The authors thank Dr Glenn A. Edwards, University ofMelbourne, for assistance with the implant of the GoretexTM

samples, and Ms Varsha Lal for assistance with ISH studies.

REFERENCES1. Ricard-Blum S. The collagen family. Cold Spring Harb Perspect

Biol 2011;3:a004978.

2. Brodsky B, Ramshaw JAM. The collagen triple-helix structure.

Matrix Biol 1997;15:545–554.

3. Goodson WH, Hunt TK. Development of a new miniature method

for the study of wound healing in human subjects. J Surg Res

1982;33:394–401.

4. White JF, Werkmeister JA, Darby IA, Bisucci T, Birk DE, Ramshaw

JAM. Collagen fibril formation in a wound healing model. J Struct

Biol 2002;137:23–30.

5. Wicke C, Halliday BJ, Scheuenstuhl H, Foree EF, Hunt TK. Exami-

nation of expanded polytetrafluoroethylene wound healing mod-

els. Wound Repair Regen 1995;3:284–291.

6. Broadbent E, Kahokehr A, Booth RJ, Thomas J, Windsor JA,

Buchanan CM, Wheeler BR, Sammour T, Hill AG. A brief relaxa-

tion intervention reduces stress and improves surgical wound

healing response: a randomised trial. Brain Behav Immun 2012;

26:212–217.

7. White JF, Werkmeister JA, Hilbert SL, Ramshaw JAM. Heart valve

collagens: cross-species comparison using immunohistological

methods. J Heart Valve Dis 2010;19:766–771.

8. Bisucci T, Hewitson TD, Darby IA. cRNA probes: Comparison of

isotopic and nonisotopic detection methods. Methods Mol Biol

2000;123:291–303.

9. Nakatsukasa H, Nagy P, Evarts RP, Hsia CC, Marsden E,

Thorgeirsson SS. Cellular distribution of transforming growth

factor-beta 1 and procollagen types I, III, and IV transcripts in car-

bon tetrachloride-induced rat liver fibrosis. J Clin Invest 1990;85:

1833–1843.

10. Werkmeister JA, Ramshaw JA, Ellender G. Characterisation of a

monoclonal antibody against native human type I collagen. Eur J

Biochem 1990;187:439–443.

11. Werkmeister JA, Ramshaw JA. Multiple antigenic determinants

on type III collagen. Biochem J 1991;274:895–898.

12. Werkmeister JA, Ramshaw JA. Monoclonal antibodies to type V

collagen for immunohistological examination of new tissue depo-

sition associated with biomaterial implants. J Histochem Cyto-

chem 1991;39:1215–1220.

13. Werkmeister JA, Tebb TA, White JF, Ramshaw JAM. Monoclonal

antibodies to type VI collagen demonstrate new tissue augmenta-

tion of a collagen-based biomaterial implant. J Histochem Cyto-

chem 1993;41:1701–1706.

14. von Recum AF, Opitz H, Wu E. Collagen types I and III at the

implant/tissue interface. J Biomed Mater Res 1993;27:757–761.

15. Shannon C, Thull R, von Recum A. Types I and III collagen in the

tissue capsules of titanium and stainless-steel implants. J Biomed

Mater Res 1997;34:401–408.

16. Werkmeister JA, White JF, Ramshaw JAM. Evaluation of the

Omniflow collagen-polymer vascular prosthesis, Med Prog Tech-

nol 1994;20:231–242.

17. Werkmeister JA, White JF, Edwards GA, Ramshaw JAM. Early

performance appraisal of the Omniflow II vascular prosthesis as

an indicator of long-term function. J Long-Term Effects Med

Implants 1995;5:1–10.

18. Werkmeister JA, Glattauer V, Tebb TA, Ramshaw JAM, Edwards

GA, Roberts G. Structural stability of long-term implants of a

collagen-based vascular prosthesis. J Long-Term Effects Med

Implants 1991;1:107–119.

19. Tsuji JM, Whitney JD, Tolentino EJ, Perrin ME, Swanson PE.

Evaluation of cellular wound healing using flow cytometry and

expanded polytetrafluroethylene implants. Wound Repair Regen

2010;18:335–340.

20. Birk DE, Br€uckner P. Collagens, suprastructures, and collagen

fibril assembly. In: Mecham RP, editor. The Extracellular Matrix:

An Overview—Biology of Extracellular Matrix. Berlin: Aspinger-

Verlag; 2011. p 77–115.

21. Wenstrup RJ, Florer JB, Brunskill EW, Bell SM, Chervoneva I, Birk

DE. Type V collagen controls the initiation of collagen fibril

assembly. J Biol Chem 2004;279:53331–53337.

22. Sun M, Chen S, Adams SM, Florer JB, Liu H, Kao WW, Wenstrup

RJ, Birk DE. Collagen V is a dominant regulator of collagen fibril-

logenesis: Dysfunctional regulation of structure and function in a

corneal-stroma-specific Col5a1-null mouse model. J Cell Sci 2011;

124:4096–4105.

23. Birk DE, Fitch JM, Babiarz JP, Doane KJ, Linsenmayer TF. Colla-

gen fibrillogenesis in vitro: Interaction of types I and V collagen

regulates fibril diameter. J Cell Sci 1990;95:649–657.

24. Marchant JK, Hahn RA, Linsenmayer TF, Birk DE. Reduction of

type V collagen using a dominant-negative strategy alters the reg-

ulation of fibrillogenesis and results in the loss of corneal-specific

fibril morphology. J Cell Biol 1996;135:1415–1426.

25. Werkmeister JA, Tebb TA, White JF, Ramshaw JAM. Collagenous

tissue formation in association with medical implants. Curr Opin

Solid State Mat Sci 2001;5:185–191.

26. Shaw LM, Olsen BR. FACIT collagens: Diverse molecular bridges

in extracellular matrices. Trends Biochem Sci 1991;16:191–194.

27. Tzortzaki EG, Tischfield JA, Sahota A, Siafakas NM, Gordon MK,

Gerecke DR. Expression of FACIT collagens XII and XIV during

bleomycin-induced pulmonary fibrosis in mice. Anat Rec A Discov

Mol Cell Evol Biol 2003;275:1073–1080.

28. Oh SP, Griffith CM, Hay ED, Olsen BR. Tissue-specific expression

of type XII collagen during mouse embryonic development. Dev

Dyn 1993;196:37–46.

29. Lethias C, Descollonges Y, Garrone R, van der Rest M. Expression

of type XIV collagen during the differentiation of fetal bovine

skin: Immunolabeling with monoclonal antibody is prominent in

morphogenetic areas. J Invest Dermatol 1993;101:92–99.

30. Castagnola P, Tavella S, Gerecke DR, Dublet B, Gordon MK, Seyer

J, Cancedda R, van der Rest M, Olsen BR. Tissue-specific expres-

sion of type XIV collagen—A member of the FACIT class of colla-

gens. Eur J Cell Biol 1992;59:340–347.

31. Berthod F, Germain L, Guignard R, Lethias C, Garrone R, Damour

O, van der Rest M, Auger FA. Differential expression of collagens

XII and XIV in human skin and in reconstructed skin. J Invest Der-

matol 1997;108:737–742.

32. Ansorge HL, Meng X, Zhang G, Veit G, Sun M, Klement JF,

Beason DP, Soslowsky LJ, Koch M, Birk DE. Type XIV Collagen

Regulates Fibrillogenesis. J Biol Chem 2009;284:8427–8438.

33. Nishiyama T, McDonough AM, Bruns RR, Burgeson RE. Type XII

and XIV collagens mediate interactions between banded collagen

fibers in vitro and may modulate extracellular matrix deformabil-

ity. J Biol Chem 1994;269:28193–28199.

34. White JF, Werkmeister JA, Edwards GA, Ramshaw JAM. Struc-

tural analysis of a collagen-polyester composite vascular prosthe-

sis. Clin Mater 1993;14:271–276.

35. Zhan Q, Burrows R, Cintron C. Localisation of type XII collagen in

normal and healing rabbit cornea by in situ hybridization. Exp

Eye Res 1995;60:551–561.

36. Massoudi D, MAlecaze F, Soler V, Butterworth J, Erraud A,

Fournie P, Koch M, Galiacy SD. NC1 long and NC3 short splice

variants of type XII collagen are overexpressed during cornel

scarring. Invest Ophthalmol Vis Sci 2012;53:7246–7256.

37. Saika S, Miyamoto T, Tanaka T, Ishida I, Ohnishi Y, Ooshima A.

Collagens XII and XIV (FACITs) in capsular opacification and in

cultured lens epithelial cells. Curr Eye Res 2001;23:463–468.

38. Tsuji JM, Whitney JD, Tolentino EJ, Perrin ME, Swanson PE.

Evaluation of cellular wound healing using flow cytometry and

expanded polytetrafluroethylene implants. Wound Repair Regen

2010;18:335–340.

39. Stamov DR, M€uller A, Wegrowski Y, Brezillon S, Franz CM. Quan-

titative analysis of type I collagen fibril regulation by lumican and

decorin using AFM. J Struct Biol 2013;183:394–403.

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | OCT 2014 VOL 102A, ISSUE 10 3555