Role of KiUer T Cells and Natural Killer 
Cell-Specific Genes in Normal 
and Pathological Immunocellular Reactions 
Several years ago Dr. Weissman's laboratory 
cloned a cDNA encoding a gene selectively ex- 
pressed in the vesicles of killer T lymphocytes and 
natural killer cells. The protein product, called 
granzyme A (GrA), is a homodimeric serine protease 
with trypsin-like specificity. GrA, other proteases, 
and perforin are released at the interface of killer 
cell-target cell junctions, causing target cell lysis. 
In the past year the laboratory demonstrated that a 
high proportion of cells in synovial fluid of patients 
with active rheumatoid arthritis are perforin and 
GrA positive. This is the first demonstration that 
cells with putative cytolytic capacity might be in- 
volved in the cytodestructive lesions of rheumatoid 
arthritis. In an examination of a number of skin le- 
sions involving lymphocytic infiltration, it was 
shown that most do not contain cells expressing 
perforin or GrA, even in lesions wherein CD8^ cells 
predominate. 
To test directly the association of killing and GrA 
positivity, the laboratory carried out a number of 
experiments wherein rat cardiac allografts pres- 
ent in mice and rats pretreated with anti-CD4 or 
cyclosporin-A (CsA) were analyzed. The anti-CD4- 
and the CsA-treated animals are tolerant but contain 
infiltrates of cells devoid of GrA and/or perforin 
message. These studies provide indirect evidence 
that perforin ± GrA^ cells are involved in cytolytic 
processes. 
Killer cells lyse target cells by two apparently in- 
dependent mechanisms: formation of membrane 
pores leading to leakage of cytoplasmic contents, 
and induced apoptosis (suicide) resulting in degra- 
dation of chromosomal DNA into nucleosomal-size 
fragments. Dr. Pierre Henkart has proposed that GrA 
is responsible for induction of apoptosis. In experi- 
ments in collaboration with Dr. Lishan Su in Dr. 
Weissman's group, transfection of GrA into mast 
cells already expressing perforin endows these cells 
with apoptosis-inducing capacity. 
Apoptosis may also occur in lymphocytes treated 
with radiation or hydrocortisone or in cell lines de- 
pendent on cytokines and then deprived of them. In 
these cases, transfection of the bcl-2 gene prevents 
apoptosis. Despite the similarity of hydrocortisone- 
induced or cytokine-removal-induced apoptosis to 
cytolytic cell-induced apoptosis. Dr. Weissman's 
laboratory found that fec/-2-transfected cells are not 
resistant to killer cell-induced apoptosis, even if 
they are resistant to apoptosis induced by cytokine 
removal. 
The laboratory examined the genomic sequence 
for the GrA gene and compared it with its mRNA 
sequences. Two cDNA sequences representing the 
products, respectively, of the 5'-most exon (EI) 
alone or that exon and its adjacent 3' exon (EI, EII) 
were represented in two RNA species. Expression of 
EI should give a hydrophobic leader to place the 
protein in the endoplasmic reticulum for transport 
to the vesicle; expression of EI, EII should result in a 
protein lacking the hydrophobic sequence, the re- 
sult of an EII termination codon followed by an 
ATG. The EII sequence lacks the hydrophobic char- 
acteristics of a signal sequence, and it is very likely 
that the protein would be expressed in the cyto- 
plasm. This EI, EII sequence can be found in a subset 
of cytolytic clones and in the thymus. 
Because secreted GrA triggers apoptosis in target 
cells, it is intriguing to consider the possibility that 
the appearance of the EI, EII form of GrA in the 
cytoplasm of killer cells and/or a subset of thymo- 
cytes might induce apoptosis in these cells. In this 
case, such expression could be involved in deter- 
mining the life span of these cells. Programmed cell 
death and programmed organismal death, by known 
genes and/or gene products have therefore become 
subjects of importance to this laboratory. Their 
study could be useful in understanding the regula- 
tion of life span as well as the mechanism by which 
the absence of positive selection and/or the pres- 
ence of negative selection cause (s) death in a subset 
of developing thymocytes. 
Dr. Weissman is also Professor of Pathology 
and Developmental Biology and (by courtesy) of 
Biology at Stanford University School of Medicine. 
Books and Chapters of Books 
Hu, M.C.-T., Holzmann, B., Neuhaus, H., and 
Weissman, I.L. 1991 . The Peyer's patch homing 
receptor: a novel member of the integrin family. 
In Vascular Adhesion Molecules (Gimbrone, 
M.A., and Cochrane, C.G., Eds.). San Diego, CA: 
Academic, pp 91-110. 
Weissman, I.L., Shih, C.-C, and Sherwood, P. 
1991. Abelson leukemia virus tumorigenesis: cel- 
lular genes that regulate growth and invasiveness. 
In Origins of Human Cancer: A Comprehensive 
Review (Brugge, J., Curran, T., Harlow, E., and 
McCormick, F., Eds.). Plainview, NY: Cold Spring 
Harbor, pp 463-471. 
Articles 
Baum, CM., Weissman, I.L., Tsukamoto, A.S., 
Buckle, A.M., and Peault, B. 1992. Isolation of a 
368 
