Connective tissue synthesis by scleroderma skin fibroblasts in cell culture.

Skin fibroblasts from subjects with scleroderma and control subjects were grown in tissue culture to compare the characteristics of connective tissue metabolism. A striking increase in soluble collagen (media hydroxyproline) was observed in eight of nine scleroderma cultures when they were compared with identically handled control cultures matched for the age and sex of the donor and the anatomic site of the donor skin. Glycoprotein content as estimated by hexosamine and sialic acid was also significantly increased in the scleroderma cultures. Estimations of protein-polysaccharide content by uronic acid determinations were low in all cultures and not significantly increased in scleroderma cultures. This report demonstrates the feasibility of using fibroblast cell cultures to study chronic rheumatic and connective tissue disorders. The initial results suggest a net increase in collagen and glycoprotein synthesis in scleroderma fibroblast cultures. The implications of an abnormality of connective tissue metabolism by skin fibroblasts propagated in vitro in the acquired disorder scleroderma are discussed.

Cross-reactivity of cell-mediated immunity between interstitial (type I) and basement membrane (type IV) collagens.

In the present study, we demonstrate delayed-type hypersensitivity (DTH) to homologous type I collagen that cross-reacts with type IV collagen. Mice immunized with native or denatured type I collagens and challenged with these same antigens or native type IV collagen develop a peak DTH response on day 7. Challenge with denatured type IV collagen or collagenase-treated type IV collagen failed to elicit DTH in type I collagen-sensitized mice. Type I collagen-sensitized spleen cells adoptively transferred DTH to types IV and I collagen to normal recipients; T cell-depleted spleen cells failed to transfer immunity. Periodate-treated type IV collagen did not elicit DTH in mice sensitized to type I collagen; however, mice sensitized with type IV collagen displayed significant DTH when challenged with periodate-treated type IV collagen. Furthermore, treatment of type IV collagen with a mixed glycosidase or alpha-glucosidase before challenge eliminated the DTH response in type I collagen-sensitized mice; beta-galactosidase treatment of type IV collagen had no effect on this response. Mice sensitized with type IV collagen, however, displayed significant DTH when challenged with these glycosidase-treated antigens. Antibodies produced to types I and IV collagen by repeated immunizations were specific for the sensitizing antigen and did not react with other connective tissue antigens. These studies indicate that a CMI response to type I collagen recognizes similar antigenic determinants on the type IV collagen molecule. These cross-reacting determinants are dependent on conformation and contain carbohydrates, particularly glucose residues.

Endothelial injury in scleroderma.

Functional and structural vascular lesions have been observed in the organs involved in scleroderma. The etiology of these vascular changes is poorly understood. The ability to isolate, characterize, and maintain endothelial cells in vitro provides a target cell population to study endothelial damage in scleroderma. The present report describes the effect of scleroderma serum on endothelial, smooth muscle, and fibroblast cell types. Sera from patients with scleroderma (31/52) and Raynaud's syndrome (11/19) contain cytotoxic activity, specific for endothelial cells, which is nondialyzable, heat-stable, and elutes with albumin on gel-filtration chromatography.

Selective upregulation of platelet-derived growth factor alpha receptors by transforming growth factor beta in scleroderma fibroblasts.

Transforming growth factor beta (TGF-beta), a multifunctional cytokine, is an indirect mitogen for human fibroblasts through platelet-derived growth factor (PDGF), particularly the A ligand-alpha receptor arm of that system. TGF-beta effects on PDGF alpha receptor expression were studied in vitro using ligand binding techniques in three human dermal fibroblast strains: newborn foreskin, adult skin, and scleroderma (systemic sclerosis, SSc). Each cell strain responded differently to TGF-beta. In newborn foreskin fibroblasts, PDGF alpha receptor number decreased in a dose-dependent manner after exposure to low concentrations of TGF-beta (0.1-1 ng/ml). Responses of normal skin fibroblasts were varied, and mean net receptor number was unchanged. Increases in PDGF alpha receptor number by TGF-beta occurred consistently with SSc fibroblasts and low concentrations of TGF-beta (0.1-1 ng/ml) were particularly stimulatory. Increased surface expression of alpha receptor subunit by TGF-beta in SSc fibroblasts correlated with increased new PDGF alpha receptor synthesis as demonstrated by radioimmunoprecipitation analysis of metabolically labeled cells and with increased steady-state levels of corresponding mRNAs. In normal adult skin fibroblasts, TGF-beta had no effect on either synthesis or mRNA expression of alpha receptor subunits. Proliferative responses to PDGF-AA after pretreatment with TGF-beta correlated positively with effects of TGF-beta on expression of alpha receptor subunit. Decreased mitogenic responses to PDGF-AA were observed in foreskin fibroblasts, small changes in responses in adult fibroblasts, and significant increases in SSc fibroblasts. Thus, costimulation with PDGF-AA and TGF-beta selectively enhanced proliferation of fibroblasts with the SSc phenotype. Immunohistochemical examination of SSc and control skin biopsies revealed the presence of PDGF-AA in SSc skin. Data obtained by ligand binding, immunoprecipitation, mRNA, and mitogenic techniques are consistent with the hypothesis that activation of the PDGF-AA ligand/alpha receptor pathway is a characteristic of the SSc fibroblast and may contribute to the expansion of fibroblasts in SSc.

Increased collagen synthesis by scleroderma skin fibroblasts in vitro: a possible defect in the regulation or activation of the scleroderma fibroblast.

Cultures of dividing skin fibroblasts from normal and sclerodermatous human skin have permitted estimations of soluble collagen concentration, net collagen accumulation, cell-doubling times, and the comparison of morphologic and ultrastructural characteristics. In vitro, the scleroderma fibroblast produces more soluble collagen, synthesizes collagen more rapidly, and fourfold more of its protein synthetic activity is directed to collagen production than in the normal skin fibroblast. Cell-doubling times and morphologic and ultrastructural observations of cells in culture have not provided clues to the nature of the biologic defect in the regulation or activation of collagen synthesis by the scleroderma fibroblast.

Activation of Hageman factor by collagen.

Purified acid-soluble and insoluble human collagen accelerated the clotting of plateletpoor plasma in silicone-treated tubes. The clot-promoting effect did not appear to be due to thromboplastic activity since the collagen preparations did not activate factor X in the presence of factor VII and calcium. Instead, collagen appeared to accelerate clotting by activating Hageman factor (factor XII) on the basis of the following findings: collagen increased the clot-promoting activity of partially purified Hageman factor but exerted no further effect in the presence of kaolin, a known activator of Hageman factor; clot-promoting eluates were obtained from collagen exposed to normal, hemophilic, or PTC-deficient plasma but not from collagen exposed to Hageman or PTA-deficient plasma. The collagen molecule itself appeared to be required for the clot-promoting activity since digestion with collagenase or thermal denaturation at pH 2.5 (about 35 degrees C) resulted in very marked reduction in clot-promoting activity. Since thermal denaturation is associated with transformation of collagen structure from triple helical to random coil form, it is suggested that the native form of collagen is essential for the ability to activate Hageman factor. Blockage of the free amino groups by treatment with nitrous acid or dinitrofluorobenzene only slightly reduced the clot-promoting activity of collagen. In contrast, since addition of cationic proteins to collagen markedly reduced pro-coagulant activity it is suggested that negatively charged sites on the collagen molecule are critical for Hageman factor activation. This suggestion is supported by the finding that pepsin treatment of collagen, which removes the predominantly negatively charged telopeptides, results in significant decrease in coagulant activity. Esterification of collagen, which neutralizes 80-90% of the free carboxyl groups, reduced coagulant activity by over 90% and it is suggested that the free carboxyl groups of glutamic and aspartic acids provide the negatively charged sites critical for Hageman factor activation.

Aggregation of platelets by collagen.

In studying some of the properties of collagen responsible for the ability to aggregate platelets it was found that thermal treatment at pH 2.5 of acid-soluble human collagen resulted in a sharp reduction in relative viscosity and platelet aggregating activity at about 35 degrees C. The reduction in viscosity is known to be associated with structural transition from triple helical to random coil form and it is postulated that the native structure of collagen is essential for its platelet aggregation effect. Blockage of the free amino groups by deamination, N-acetylation, or treatment with dinitrofluorobenzene resulted in over 90% reduction in platelet aggregating activity. Addition of cationic proteins to collagen, removal of the negatively charged telopeptides by treatment with pepsin, or acetylation of the free carboxyl groups did not significantly affect the platelet aggregating activity of collagen. On the basis of these findings it is suggested that the free amino groups and specifically the epsilon amino groups of lysine are critical for the platelet aggregating activity of collagen whereas the carboxyl groups are of relatively little importance.

Mononuclear cell modulation of connective tissue function: suppression of fibroblast growth by stimulation of endogenous prostaglandin production.

The role of immune cell products in modulating connective tissue metabolism was investigated. Supernates of both unstimulated and phytohemagglutinin-stimulated human mononuclear cell cultures suppressed fibroblast proliferation (up to 90%) and concomintantly stimulated fibroblast prostaglandin E(PGE) synthesis (20- to 70-fold). The growth suppression was, at least in part, a secondary result of the increased fibroblast PGE synthesis; growth suppression (a) paralled the increased fibroblast PGE synthesis, (b) was reversed by addition of inhibitors of prostaglandin synthesis (indomethacin, meclofenamate, and eicostaetraynoic acid), and (c) was reproduced by addition of exogenous PGE2 to fibroblast cultures. The prostaglandin-stimulatory, growth-suppressive activity was a product of non-T-lymphocyte, adherent cells and was present within 6 h of mononuclear cell culture. The activity was heat (56 degrees C) and trypsin sensitive, nondialyzable, and appeared in the 12,000-20,000 mol wt fractions by Sephadex G-100 chromatography. The activity in supernates of mononuclear cell cultures was removed by incubation with fibroblasts but not by similar incubation with peripheral blood mononuclear cells. Mononuclear cells release a factor(s) which modulates fibroblast proliferation by altering prostaglandin metabolism.

Skin capillary blood flow in scleroderma.

Skin blood flow was measured by the clearance of radioactive xenon ((133)Xe) injected intracutaneously in eight patients with scleroderma and nine control subjects under conditions of controlled temperature and humidity. Scleroderma patients, on being cooled 1 hr at 18 degrees C, had a rate constant of (133)Xe clearance from the dorsal finger skin which was 0.04 +/-0.07 min(-1) (mean +/-SD), compared with 0.23 +/-0.15 min(-1) in normal subjects (P < 0.005). The corresponding mean cutaneous blood flows were 2.9 ml/100 g per min in the scleroderma patients and 16.4 ml/100 g per min in normal subjects. After reflex warming by waterbath, clearance was similar in the two groups (0.33 +/-0.1 vs. 0.40 +/-0.09); these data suggest that diminished clearance in scleroderma patients on cooling resulted, at least in part, from functional or reversible interruption of the circulation. The skin temperatures of scleroderma patients after reflex warming remained lower than those of normal subjects, despite similar increases in sublingual temperatures. The dissociation of (133)Xe clearance and skin temperature in scleroderma patients (i.e. subnormal skin temperatures with normal (133)Xe clearance after reflex warming) suggests either abnormal thermal properties of scleroderma skin or selective vasoconstriction of the vessels which regulate heat exchange. The demonstrated interruption of the capillary circulation on cooling of the skin in patients with scleroderma may be important in the pathogenesis of this disorder. After oral pretreatment with guanethidine, five patients with scleroderma had increased (133)Xe clearance and calculated blood flow on cooling, rising to normal in three of these patients. The potential of this technique for the quantitative sequential evaluation of skin blood flow in subjects with scleroderma and for the evaluation of empirical therapy is suggested.

Effects of neutrophil elastase and other proteases on porcine aortic endothelial prostaglandin I2 production, adenine nucleotide release, and responses to vasoactive agents.

The effects of neutrophil elastase on endothelial prostacyclin (PGI2) production, nucleotide release, and responsiveness to vasoactive agents were compared with the effects of cathepsin G (the other major neutral protease of neutrophils), pancreatic elastase, trypsin, chymotrypsin, and thrombin. PGI2 production by pig aortic endothelial cells cultured on microcarrier beads and perfused in columns was stimulated in a dose-dependent manner by trypsin, chymotrypsin, and cathepsin G (1-100 micrograms/ml for 3 min). Thrombin, while active at low concentrations (0.1-10 National Institutes of Health U/ml), induced smaller responses. Neutrophil and pancreatic elastase had little or no effect on PGI2 production. Dose-dependent, selective release of adenine nucleotides was induced by neutrophil elastase (3-30 micrograms/ml). The other proteases were much less active; for example, trypsin (100 micrograms/ml) induced a response only approximately 5% as great as did 30 micrograms/ml neutrophil elastase. After exposure to 30 micrograms/ml neutrophil elastase, cells did not exhibit the characteristic burst of PGI2 production in response to extracellular ATP; responsiveness gradually returned after 40-120 min. This effect was not seen with the other proteases. Elastase partly inhibited responses to bradykinin and had no effect on PGI2 production that was stimulated by ionophore A23187. There was no evidence of cytotoxicity, as measured by release of lactate dehydrogenase. Neutrophil degranulation can generate concentrations of elastase and cathepsin G comparable with those tested in the present study, and the effects of these enzymes on endothelial function lead us to suggest that they may play a role in vasoregulation and vascular pathology.

Elevated expression of c-myc proto-oncogene in scleroderma fibroblasts.

Scleroderma is a connective tissue disease characterized by the overproduction of extracellular matrix components. The mechanisms of fibrosis may involve increased fibroblast proliferation in the scleroderma lesion due to the presence of cells with abnormal growth properties, in addition to the well-known over-production of several matrix components. The c-myc proto-oncogene has been implicated in dysregulation of cell growth in neoplastic cells and as an essential element of the response to growth factors in normal cells. Therefore, to investigate the molecular basis of growth in scleroderma, we compared expression of c-myc gene in scleroderma and control cells. In this report, we show that under low serum conditions (1% serum), scleroderma fibroblasts express 2.5-3 higher level of c-myc message. Moreover, stimulation of c-myc after addition of fresh 10% serum is blunted in scleroderma compared to control cells. Observed c-myc expression in scleroderma is similar to c-myc expression in transformed cells. In addition, there is also increased proliferation of scleroderma cells in 1% serum as measured directly by a nuclear label assay. These data suggest the presence of fibroblasts with abnormal growth properties in the scleroderma lesion.

Pathogenesis of fibrosis: type 1 collagen and the skin.

This review on the pathogenesis of fibrosis emphasizes the similarities between tissue repair, a tightly regulated salutary biological response, and fibrosis, an unregulated pathological process. It focuses on the transcriptional regulation of type I collagen, the role of cytokines in fibroblast activation, integrins as examples of cell-matrix signaling pathways, and the heterogeneity of fibroblast populations as factors contributing to fibrosis. Tissue remodeling and the role of matrix metalloproteinases and metalloproteinase inhibitors are mentioned briefly. The capacity of extracellular matrix to modulate cellular function is a recurring theme.

Pathogenesis of scleroderma (systemic sclerosis).

Increasing interest in the vascular features of scleroderma has led to the hypothesis that the blood vessel is the major target tissue and that the endothelial cell is the principal cell target. Useful observations stemming from the vascular hypothesis include the use of microvascular abnormalities in the early detection of the patient destined to develop classical scleroderma, the discovery of a serum protease selectively cytotoxic to endothelial cells, and the study of a serum mitogenic activity for fibroblasts in scleroderma patients. Immune events related to the vascular lesions are under active study but have not as yet provided a unique immunological lesion in scleroderma patients. The possibility that immunity to basement membrane (type IV) collagen may be selective for scleroderma patients deserves further study. Persistent immunity to endothelial basement membrane structures would provide a basis for continued endothelial injury. Techniques to quantify endothelial injury are useful to assess activity of the vascular lesions and to monitor therapies designed to block further vascular injury. The definition of pre-fibrotic vascular lesions may have future therapeutic and preventive implications for scleroderma.

A possible role for transforming growth factor-beta in systemic sclerosis.

The cause of systemic sclerosis remains unknown, but cellular and molecular mechanisms possibly responsible for the characteristic clinical manifestations of fibrosis and vascular damage (Raynaud's phenomenon, telangiectasis, digital infection, and renal arteriopathy) are becoming understood in greater detail. One possibly important cytokine is transforming growth factor-beta (TGF-beta); its involvement is reviewed here. With regard to vascular lesions, TGF-beta has variably been shown to inhibit endothelial cell growth in vitro but to promote angiogenesis in vivo, a paradox that remains unresolved. Nonetheless, an injurious activity of TGF-beta on microvascular endothelial cells could help to explain the intimal proliferation and microvascular obliteration seen. Whether as a result of or as a cause of endothelial cell damage, platelet activation has been well documented in systemic sclerosis and the platelet alpha granule pool contains a large quantity of TGF-beta. TGF-beta is also produced by activated macrophages and T cells, both of which are known to occur within systemic sclerosis lesions. An important effect of TGF-beta is its stimulation of fibroblast collagen and fibronectin synthesis and their deposition into the extracellular matrix. Stimulation by TGF-beta may therefore account for the fibrosis seen in the dermis and in the internal organs. Direct evidence of TGF-beta involvement in systemic sclerosis is scanty, and awaits discovery of either an abnormal expression of or response to TGF-beta. The biologic effects of TGF-beta appear to be regulated at the level of activation from a latent polypeptide precursor form. Descriptions of the importance of this cytokine in pathologic conditions will need to account for this activation and its regulation. Nonetheless, the physiologic effects so far attributed to TGF-beta make its involvement in systemic sclerosis an attractive possibility to explain some of the manifestations of this enigmatic disease.

Different effects of basic fibroblast growth factor and transforming growth factor-beta on the two platelet-derived growth factor receptors' expression in scleroderma and healthy human dermal fibroblasts.

Previous studies have demonstrated that platelet-derived growth factor (PDGF) alpha receptor expression is up-regulated by transforming growth factor-beta 1 (TGF-beta 1) in scleroderma dermal fibroblasts, but not in healthy control fibroblasts. We asked whether this selective effect in scleroderma cells was TGF-beta 1-specific or a general response by studying responses to other growth factors. In this study, we compared the expression of alpha and beta PDGF receptor subunits (mRNA and protein levels) in these two cell types in response to basic fibroblast growth factor (bFGF) and TGF-beta 1. bFGF coordinately stimulated mRNA levels of alpha and beta receptor subunits in healthy fibroblasts, but did not change PDGF receptor expression in scleroderma fibroblasts. Conversely, and in agreement with previous observations, TGF-beta 1 induced PDGF alpha receptor expression in scleroderma fibroblasts, but not in healthy fibroblasts. PDGF beta receptor mRNA levels were induced to similar degrees by TGF-beta 1 in both cell types. PDGF alpha receptor protein levels correlated directly with mRNA levels, induced by bFGF only in healthy fibroblasts and by TGF-beta 1 only in scleroderma fibroblasts. However, PDGF beta receptor protein levels were not altered by either growth factor in either cell type. Thus, the activated state of scleroderma fibroblasts does not include receptor-signaling pathways to bFGF. This distinct pattern of expression of PDGF alpha receptors in scleroderma fibroblasts suggests a possible role for the coordinately expressed PDGF AA ligand/alpha receptor system in the development of fibrosis.

Differential modulation of bFGF receptors by TGF-beta in adult skin, scleroderma skin, and newborn foreskin fibroblasts.

Effects of transforming growth factor beta (TGF-beta) on proliferative responses to basic fibroblast growth factor (bFGF) were studied in human diploid fibroblasts cell strains derived from three different sources: adult skin, scleroderma, and newborn foreskin. All three types of cell strains were similarly responsive to TGF-beta, whereas adult skin fibroblasts were significantly more responsive to bFGF. Incubation of cells with TGF-beta prior to bFGF addition substantially increased responsiveness of adult skin fibroblasts to this latter cytokine, slightly increased that of scleroderma fibroblasts, and decreased that of foreskin fibroblasts. Modulation of bFGF receptors by TGF-beta correlated positively with these mitogenic effects. Adult skin fibroblasts showed increases of both high- and low-affinity receptors and scleroderma fibroblasts showed small increases of high-affinity receptors only, whereas foreskin fibroblasts showed no changes. Heparitinase treatment of adult skin fibroblasts during TGF-beta pre-incubation resulted in reduced bFGF binding to low-affinity receptors and reduced mitogenic response to bFGF, suggesting that the TGF-beta-stimulated increase of low-affinity receptors in these cells contributes to the observed enhanced mitogenic effects of bFGF. Abnormal responses of scleroderma fibroblasts to TGF-beta/bFGF stimulation, particularly failure to synthesize low-affinity receptors in response to TGF-beta, adds a new characteristic to the fibrotic phenotype of scleroderma fibroblasts.

IL-4 upregulates tenascin synthesis in scleroderma and healthy skin fibroblasts.

Tenascin (TN), a large extracellular matrix glycoprotein, is transiently expressed during embryonic development, but is absent from most normal adult tissues. TN is reexpressed, however, in healing wounds, in the stroma of some tumors, and in fibrotic diseases such as systemic sclerosis (SSc) and rheumatoid arthritis. To clarify the mechanisms regulating TN expression, we studied the effects of selected cytokines (PDGF, bFGF, TGF-beta, IL-1, IL-4, IL-6, IFN-gamma, and TNF-alpha) found in fibrotic tissue on TN expression by dermal fibroblasts. IL-4 strongly induced TN protein levels (up to 10-fold over the basal level), whereas PDGF and bFGF were less potent inducers of TN than IL-4. All other cytokines tested, including TGF-alpha1, did not stimulate TN synthesis. IL-4 also increased TN mRNA expression, and this effect was blocked by actinomycin D. Cycloheximide increased basal TN mRNA expression and induced TN mRNA in IL-4-treated fibroblasts, suggesting that repressor protein(s) may regulate transcription of the TN gene. Although no differences in constitutive TN expression or effects of cytokines on TN expression were observed between SSc and healthy fibroblasts, these data are consistent with the observations that high levels of both IL-4 and TN are present in the affected skin of patients with SSc. These results suggest that the high level of TN found in the affected tissue of patients with SSc results from the high level of IL-4 present.

The management of renal scleroderma: experience with dialysis, nephrectomy and transplantation.

In 25 of 100 patients with scleroderma seen over a five year period irreversible renal failure developed; renal support was instituted in 17. Ten of 17 received peritoneal or hemodialysis, one survived. The remaining seven received hemodialysis plus nephrectomy; three survived. Two of these three underwent renal transplantation; one survived. This experience is presented to encourage improvement of these and other technics to increase the survival rate in the otherwise uniformly fatal renal failure associated with scleroderma (systemic sclerosis).

Skin capillary abnormalities as indicators of organ involvement in scleroderma (systemic sclerosis), Raynaud's syndrome and dermatomyositis.

Forty-four study patients with scleroderma (systemic sclerosis) (28 patients), Raynaud's syndrome (13 patients) or dermatomyositis (three patients) were observed for skin capillary abnormalities by widefield microscopy and compared with three control groups of 20 subjects each: (1) patients with other rheumatic disease, (2) hospitalized patients with nonrheumatic conditions, and (3) healthy volunteers. The distinctive microvascular pattern (dilated and distorted capillary loops alternating with avascular areas) previously reported in scleroderma and dermatomyositis was observed almost exclusively in the study patients. The severity of capillary abnormalities varied among the diagnostic subgroups, and a positive correlation was found between the degree and extent of abnormal microvascular patterns and multisystem involvement. On this basis, widefield nailfold capillary observations are proposed as a simple, inexpensive, reproducible technic for making an improved early diagnosis and predicting multisystem involvement in scleroderma, Raynaud's syndrome and dermatomyositis, presently a group of loosely associated and overlapping connective tissue disorders which often defy early and precise diagnosis.

A prospective study of Raynaud phenomenon and early connective tissue disease. A five-year report.

To define features of patients with Raynaud phenomenon that predict the evolution of connective tissue disorders, a prospective study was initiated. Patients with history or physical evidence of Raynaud phenomenon without causal or associated disorders underwent initial multisystem evaluation. Scleroderma (systemic sclerosis) of less than five years' duration was included for comparison. Patients were classified as having Raynaud phenomenon "only," undifferentiated connective tissue syndrome or scleroderma. Nailfold capillary microscopy was performed, and patterns were scored blindly from coded photographs. Of 91 patients with Raynaud phenomenon entered (Raynaud phenomenon only, n = 49; undifferentiated connective tissue syndrome, n = 22; scleroderma, n = 20), abnormal "scleroderma pattern" capillaries were noted in seven of 49, 19 of 22 and 19 of 20, respectively (p less than 0.005). Of 39 patients with Raynaud phenomenon only followed (mean duration, 23.7 months) undifferentiated connective tissue syndrome developed in three. Of 17 patients with undifferentiated connective tissue syndrome followed (mean duration, 27.6 months), six patients had transitions (four scleroderma; one scleroderma-systemic lupus erythematosus overlap; one SLE). Nailfold capillary abnormalities best identified transitional patients in both groups (eight of nine) and were more sensitive than ANA (six of nine), presence of digital ulcers (four of nine) or decreased esophageal motility (two of nine). This prospective study documents a useful role for capillary examinations in evaluating Raynaud phenomenon. Abnormal capillaries indicate an increased risk for connective tissue disease; normal capillaries favor idiopathic Raynaud phenomenon.