By the use of sucrose gelatin veronal buffer (SGVB), a simple screening test was developed by us to detect sera with low complement activity, including C9-deficient sera. Using this screening test, we were able to identify sera with low complement activity including C9-deficient sera among a large number of samples. Further examinations, estimation of the protein concentration of C9, C4, C3, etc., enabled classification of serum with low complement activity into C9-deficient serum, serum deficient in the other components, and serum with low complement activity caused by non-specific activation of complement through the classical pathway by low temperature in vitro. Among 145,640 sera from Osaka donors, 138 sera were found to be deficient in C9 by these methods. The whole complement activity (CH50) of the 138 sera was 13.1 +/- 3.0 U/ml. The C9 protein in these sera was undetectable, not only by the single radial immunodiffusion method, but also by the sensitive ELISA method. C9 activities in these sera were less than 0.1% of the level in pooled normal human serum. These findings and the family studies revealed that 138 blood donors unquestionably had a hereditary C9 deficiency. The incidence of C9 deficiency among Osaka donors was calculated to be 0.095%.

Among sera from 145,640 healthy blood donors in Osaka, 16 were found to have abnormalities in late-acting complement components other than C9. It was found that of these 16 sera, 2 were deficient in C5, 4 in C6, 6 in C7 and 4 in the C8 α-γ-subunit. The incidence of deficiency of each component among the Osaka blood donors was calculated as follows: C5 deficiency, 0.0014%; C6 deficiency, 0.0027%; C7 deficiency, 0.0041 %; C8 α-γ-subunit deficiency, 0.0027%. We confirmed that 13 donors were healthy and 12 had no past history related to a complement component deficiency. From these results, not only C9 deficiency but also deficiencies of the other late-acting complement components were found among the healthy blood donors, but no early-acting component deficiencies were noted.

From 92,686 sera sent from hospitals throughout Japan to the Special Reference Laboratories, for CH50 assay, we were able to classify 80 patients as C9-deficient using a sensitive screening test, as well as hemolytic and immunochemical C9 assays. The incidence of C9 deficiency was determined to be 0.086%, and there were no distinct differences among the eight areas of Japan tested. Serum CH50 levels of these C9-deficient patients varied widely (9.4–63.8 U/ml), and exhibited a higher value (average: 34.1 U/ml) than that of healthy C9-deficient individuals, probably due to elevated C3, C4, and C5 levels. These patients suffered from a variety of autoimmune, renal, and infectious diseases, which, however, are thought to be only incidentally associated with C9 deficiency.

We observed two sisters with lupus-like syndrome with homozygous C3 deficiencies. A 19-year-old woman and her 15-year-old sister developed malar rash, arthralgia, and photosensitivity, but antinuclear antibodies and LE cell preparations were negative. The older sister experienced recurrent bronchitis in her childhood, but the younger sister had no recurrent infections. Serum C3 was not detected immunochemically in either sister, and total complement activity and C3 hemolytic activity were extremely low.