Anticentromere antibodies (ACA) are immunological markers for the subset of systemic scleroderma with the symptoms calcinosis cutis, Raynaud's phenomenon, esophageal dysfunction, sclerodactyly, and telangiectasia (CREST). In Western blotting, some ACA-positive sera also recognize a doublet of 23 kDa (p23) and 25 kDa (p25) in addition to centromere protein antigens A (17 kDa), B (80 kDa), and C (140 kDa). Two forms of p25 have been shown to be human homologues of Drosophila heterochromatin-associated protein HP1. One form of p25 (p25beta) which was recently cloned in this laboratory was used to evaluate anti-p25beta antibody response in scleroderma sera. Of 318 scleroderma sera 42 had ACA (13.2%), and 16 of the 42 sera (38%) had anti-p25beta antibodies. On the other hand, 5 of 276 ACA-negative sera (1.8%) showed anti-p25beta antibody response, demonstrating that anti-p25beta antibody is significantly associated with the ACA response (P < 10[-8]). Clinically the anti-p25beta response was significantly associated with the CREST syndrome. Fourteen (36.8%) of 38 CREST patients compared to seven (2.5%) of 280 patients with other forms of scleroderma were anti-p25beta antibody positive (P < 10[-8]). The 14 CREST patients with anti-p25beta antibodies had significantly more interstitial lung disease than those without anti-p25beta antibodies (P < 0.003). There was also a tendency to increased liver involvement. Two dominant autoepitopes in p25beta were determined by Western blotting using p25beta recombinant fragments. In immunofluorescence C-terminal specific antibodies showed staining of heterochromatin, but N-terminal specific antibodies showed no staining. Interestingly, the majority of sera reacted preferentially with one or the other of the two dominant autoepitopes.