Volume 387, Issue 6

The human kallikrein locus on chromosome 19q13.3–13.4 contains kallikrein 1 – the tissue kallikrein – and 14 related serine proteases. Recent investigations into their function and evolution have indicated that the present nomenclature for these proteins is inadequate or insufficient. Here we present a new nomenclature in which proteins without proven kininogenase activity are denoted kallikrein-related peptidase. Names are also given to the unique rodent proteins that are closely related to kallikrein 1.

Human tissue kallikreins (hKs) are attracting increased attention owing to their association with various forms of cancer and other diseases. Human tissue kallikrein genes represent the largest contiguous group of proteases within the human genome. There are many areas of kallikrein research that need to be further explored, including their tissue expression patterns, their regulation, identification of specific substrates, their participation in proteolytic cascades, and their clinical applicability as cancer biomarkers and therapeutic targets. In this review, we briefly describe the current status of kallikrein research and identify future avenues that will enhance our understanding of their function and involvement in human diseases.

We have studied the immunohistochemical expression (IE) of eight non-tissue-specific human kallikreins (hKs) (hK5, 6, 7, 10, 11, 12, 13, and 14) in different normal tissues. The IE was always cytoplasmic, showing a characteristic pattern in some tissues. Comparison of the IE of all hKs studied in the different tissues revealed no major differences, suggesting that they share a common mode of regulation. Furthermore, hKs were immunohistochemically revealed in a variety of tissues, indicating that no protein is tissue-specific (except for hK2 and hK3, which have tissue-restricted expression). In general, our results correspond well with data from RT-PCR and ELISA assays. Glandular epithelia constitute the main kallikrein IE sites, and the staining in their secretions confirms that these proteases are secreted. A variety of other tissues express the proteins as well. We have also immunohistochemically evaluated all the above hKs in several malignant tissues. Tumors arising from tissues expressing kallikreins tested positive. Corresponding to the IE in normal glandular tissues, most hKs were expressed in adenocarcinomas. The prognostic value of several hKs was studied in series of prostate, renal cell, colon and urothelial carcinomas.

Tissue kallikrein (hK1) cleaves low-molecular-weight kininogen to produce kinin peptide, which binds to kinin receptors and triggers a wide spectrum of biological effects. Tissue kallikrein levels are reduced in humans and in animal models with hypertension, cardiovascular and renal diseases. Transgenic mice or rats over-expressing human tissue kallikrein or kinin B2 receptor are permanently hypotensive, and somatic kallikrein gene delivery reduces blood pressure in several hypertensive rat models. Moreover, kallikrein gene delivery or kallikrein protein infusion can directly improve cardiac, renal and neurological function without blood pressure reduction. Kallikrein has pleiotropic effects in inhibiting apoptosis, inflammation, proliferation, hypertrophy and fibrosis, and promoting angiogenesis and neurogenesis in different experimental animal models. Kallikrein's effects can be blocked by kinin B2 receptor antagonists. Mechanistically, tissue kallikrein/kinin leads to increased nitric oxide levels and Akt activation, and reduced reactive oxygen species formation, TGF-β1 expression, MAPK and nuclear factor-κB activation. Our studies indicate that tissue kallikrein, through the kinin B2 receptor and nitric oxide formation, can protect against oxidative damage in cardiovascular and renal diseases and ischemic stroke. These novel findings suggest that kallikrein/kinin may serve as new drug targets for the prevention and treatment of heart failure, renal disease and stroke in humans.

Serine proteinases, like trypsin, can play a hormone-like role by triggering signal transduction pathways in target cells. In many respects these hormone-like actions of proteinases can now be understood in terms of the pharmacodynamics of the G protein-coupled ‘receptor’ responsible for the cellular actions of thrombin (proteinase-activated receptor-1, or PAR 1 ). PAR 1 , like the other three members of this receptor family (PAR 2 , PAR 3 and PAR 4 ), has a unique mechanism of activation involving the proteolytic unmasking of an N-terminally tethered sequence that can activate the receptor. The selective activation of each PAR by short synthetic peptides representing these sequences has demonstrated that PAR 1 , PAR 2 and PAR 4 play important roles in regulating physiological responses ranging from vasoregulation and cell growth to inflammation and nociception. We hypothesise that the tissue kallikreins may regulate signal transduction via the PARs. Although PARs can account for many of their biological actions, kallikreins may also cause effects by mechanisms not involving the PARs. For instance, trypsin activates the insulin receptor and thrombin can act via a mechanism involving its non-catalytic domains. Based on the data we summarise, we propose that the kallikreins, like thrombin and trypsin, must now be considered as important ‘hormonal’ regulators of tissue function.

Kallikreins have been implicated in carcinogenesis and are promising biomarkers for the diagnosis and follow-up of various cancers. To evaluate the functions and clinical interest of kallikreins, it is important to be able to produce them as recombinant proteins. Here we summarize the various strategies used to produce kallikreins, emphasizing their advantages and limitations. We also describe an approach to achieve high-level production of kallikreins, such as hK6, with correct folding and activity, combining an expression system with targeted transgene integration and an efficient cultivation device to increase yield, the CELLine bioreactor. This novel method for recombinant kallikrein production will be useful to study their bio-pathological functions and to develop anti-bodies.

The kallikrein-related peptidase ( KLK ) gene family of 15 serine proteases encodes many proteins, including prostate specific antigen (PSA or KLK3), that are well described and/or are potential biomarkers for hormone-related cancers. Variant mRNA transcripts produced by alternative splicing, polyadenylation or AUG sites, or intron retention have been found for each of the KLK genes. The predicted protein for many of these alternative transcripts is different from that of the classical kallikrein-related peptidases and would not be an active serine protease. The majority of these novel protein isoforms have not been studied in vivo . The possible function(s) of the variant transcripts/protein isoforms and potential roles that they may play in hormone-related cancers are still unknown and are the focus of this short review.

Several members of the kallikrein-related peptidase family of serine proteases have proteolytic activities that may affect cancer progression; however, the in vivo significance of these activities remains uncertain. We have demonstrated that expression of PSA or KLK4, but not KLK2, in PC-3 prostate cancer cells changed the cellular morphology from epithelial to spindle-shaped, markedly reduced E-cadherin expression, increased vimentin expression and increased cellular migration. These changes are indicative of an epithelial to mesenchymal transition (EMT), a process important in embryonic development and cancer progression. The potential novel role of kallikrein-related peptidases in this process is the focus of this brief review.

Our laboratory is involved in identifying genes that can be used as early diagnostic or prognostic markers in breast cancer. We previously identified a gene ( NES1 ) that is expressed in normal but not in transformed mammary epithelial cells (MECs). NES1 is located on chromosome 19q13.4 within the kallikrein locus and thus was designated as human kallikrein 10 ( hK10 ), although we have been unable to detect any protease activity. Importantly, hK10 expression is decreased in a majority of breast cancer cell lines. Transfection of hK10 into hK10 -negative breast cancer cells reduces the tumorigenicity. Using methylation-specific PCR and subsequent sequencing, we demonstrate a strong correlation between hypermethylation of hK10 and loss of mRNA expression. Further analysis showed that essentially 100% of normal breast specimens had hK10 expression, whereas 46% of ductal carcinoma in situ (DCIS) and the majority of infiltrating ductal carcinoma (IDC) samples lacked the hK10  mRNA. Importantly, hK10 -negative DCIS diagnosed at the time of biopsy were subsequently diagnosed as IDC at the time of definitive surgery. It has been shown that hK10 protein expression is regulated by steroids. In addition to breast cancers, hK10 is downregulated in cervical cancer, prostate cancer and acute lymphocytic leukemia, whereas it is upregulated in ovarian cancers. These results point to the paradoxical role of hK10 in human cancers and underscore the importance of further studies of this kallikrein.

Human kallikrein 8 (hK8), whose gene was originally cloned as the human ortholog of a mouse brain protease, is known to be associated with diseases such as ovarian cancer and Alzheimer's disease. Recombinant human pro-kallikrein 8 was activated with lysyl endopeptidase-conjugated beads. Amino-terminal sequencing of the activated enzyme demonstrated the cleavage of a 9-aa propeptide from the pro-enzyme. The substrate specificity of activated hK8 was characterized using synthetic fluorescent substrates. hK8 showed trypsin-like specificity, as predicted from sequence analysis and enzymatic characterization of the mouse ortholog. All synthetic substrates tested containing either arginine or lysine at P1 position were cleaved by hK8. The highest k cat / K m value of 20×10 3 M -1 s -1 was observed with Boc-Val-Pro-Arg-7-amido-4-methylcoumarin. The activity of hK8 was inhibited by antipain, chymostatin, and leupeptin. The concentration for 50% inhibition by the best inhibitor, antipain, was 0.46 μM. The effect of different metal ions on the enzyme activity was analyzed. Whereas Na + had no effect on hK8 activity, Ni 2+ and Zn 2+ decreased the activity and Ca 2+ , Mg 2+ , and K + had a stimulatory effect. Ca 2+ was the best activator, with an optimal concentration of approximately 10 μM.

Human tissue kallikreins (genes, KLK s; proteins, hKs) are a subgroup of hormonally regulated serine proteases. Two tissue kallikreins, namely hK2 and hK3 (prostate-specific antigen, PSA), are currently used as serological biomarkers of prostate cancer. Human tissue kallikrein 9 ( KLK9 ) is a newly identified member of the tissue kallikrein gene family. Recent reports have indicated that KLK9 mRNA is differentially expressed in ovarian and breast cancer and has prognostic value. Here, we report the production of recombinant hK9 (classic form) using prokaryotic and mammalian cells and the generation of polyclonal antibodies. Total testis tissue mRNA was reverse-transcribed to cDNA, amplified, cloned into a pET/200 TOPO plasmid vector, and transformed into E. coli cells. hK9 was purified and used as an immunogen to generate polyclonal antibodies. Full-length KLK9 cDNA was also cloned in the vector pcDNA3.1 and was expressed in CHO cells. The identity of hK9 was confirmed by mass spectrometry. hK9 rabbit antiserum displayed no cross-reactivity with other tissue kallikreins and could specifically recognize E. coli - and CHO-derived hK9 on Western blots. hK9 was mainly detected in testis and seminal vesicles by Western blotting. The reagents generated here will help to define the physiological role of this tissue kallikrein and its involvement in human disease.

Human kallikrein 10 (hK10) protein is expressed in normal breast but is significantly downregulated in a majority of invasive breast cancers. Thus, understanding how hK10 expression is regulated is of substantial significance. In this study, we analyzed the promoter region of hK10 using a website software (TRANSFAC 3.0), which predicted three possible retinoic acid response elements (RAREs), RARE1 at -1041 ( TGACCT CG TGATCC ), RARE2 at -859 ( TGACCT C CTATGA ) and RARE3 at -765 ( TGACCT C CTGTGA ), each with a half-site of a canonical sequence (TGACCT; reverse complement AGGTCA). Using electrophoretic mobility shift assays and nucleotide competition analysis, as well as chromatin immunoprecipitation of the native hK10 promoter, we demonstrated specific binding of RXR only to RARE1. The functional importance of RARE in the hK10 promoter was demonstrated by retinoid induction of hk10 promoter-reporters; furthermore, mutation of RARE1 but not of RARE2 or RARE3 abolished the induction of the reporter. Finally, we demonstrated the induction of hK10 mRNA and protein expression upon retinoid treatment of cells. In view of the correlation of the downregulation of hK10 mRNA and protein with breast cancer progression, these findings suggest a potential approach to restore hK10 expression in cancer patients.

Human kallikrein 4 (hK4) is a member of the expanded family of human kallikreins, a group of 15 secreted proteases. While this protein has been associated with ovarian and prostate cancer prognosis, only limited functional information exists. Therefore, we have undertaken an investigation of its enzymatic properties regarding substrate preference, degradation of extracellular matrix proteins, and its inhibition by various inhibitors. We successfully expressed and purified active recombinant hK4 from supernatants of the Pichia pastoris expression system. This enzyme seems to cleave more efficiently after Arg compared to Lys at the P1 position and exhibits modest specificity for amino acids at positions P2 and P3. hK4 forms complexes with α 1 -antitrypsin, α 2 -antiplasmin and α 2 -macroglobulin. The protease mediates limited degradation of extracellular matrix proteins such as collagen I and IV, and more efficient degradation of the α-chain of fibrinogen. The cleavage of extracellular matrix proteins by hK4 suggests that this enzyme may play a role in tissue remodeling and cancer metastasis.

We have previously presented evidence that two human kallikrein-related peptidases, KLK5 (hK5, stratum corneum tryptic enzyme, SCTE) and KLK7 (hK7, stratum corneum chymotryptic enzyme, SCCE), which are abundant in the stratum corneum, may be involved in desquamation. Since we had noted that not all trypsin-like activity in the plantar stratum corneum could be ascribed to KLK5, we set out to identify other skin proteases with similar primary substrate specificity. Here we describe purification of a protease identified as KLK14 from plantar stratum corneum, and show that this enzyme may be responsible for as much as 50% of the total trypsin-like activity in this tissue, measured as activity towards a chromogenic substrate cleaved by a wide variety of enzymes with trypsin-like specificity. This was in spite of very low levels of KLK14 protein compared to KLK5 and KLK7. KLK14 could be detected by immunoblotting in normal superficial stratum corneum of all individuals examined. The majority of KLK14 in the plantar stratum corneum is present in its catalytically active form. KLK14 could be immunohistochemically detected in sweat ducts, preferentially in the intraepidermal parts (the acrosyringium), and in sweat glands. The role played by this very efficient protease under normal and disease conditions in the skin remains to be elucidated.

Prostate-specific antigen (PSA) is a widely used marker for prostate cancer. The utility of PSA tests is limited by their inability to differentiate prostate cancer from non-malignant conditions such as benign prostatic hyperplasia and prostatitis. In circulation, PSA occurs in various complexed and free forms, and specific determination of some of these can be used to improve the diagnostic accuracy of PSA tests. We have previously identified peptides that specifically bind to enzymatically active PSA and using such a peptide we have developed an immunopeptidometric assay for this form of PSA. However, the sensitivity of that assay is too low to measure active PSA at clinically important levels. Recently a novel sensitive immunoassay for analysis of proteins, termed the proximity ligation assay, has been established. Here we describe a sensitive implementation of the proximity ligation assay, which utilizes a PSA-binding peptide and antibody as probes to detect active PSA. The assay has a sensitivity of 0.07 μg/l, which is approximately ten-fold lower than that of our previous assay. It does not cross-react with inactive proPSA or the highly similar kallikrein hK2. Our results show that a highly sensitive immunopeptidometric assay can be developed using proximity ligation. This principle should facilitate establishment of specific assays for active forms of other proteases.

Human kallikrein 6 ( KLK6 ) was identified based on its transient upregulation in a primary breast tumor and its subsequent silencing in a metastatic tumor from the same patient. The molecular mechanism(s) underlying the deregulated expression of KLK6 during cancer progression are currently unknown. Here, we provide evidence that aberrant expression of KLK6 is regulated at the level of transcription by multiple cooperating mechanisms. KLK6 can be reactivated in non-expressing breast cancer cells by treatment with 5-aza-2′-deoxycytidine (5-aza-dC), a compound causing DNA demethylation. Trichostatin A (TSA), an inhibitor of histone deacetylases, resulted in moderate induction of KLK6 only in MDA-MB-231 cells. However, combined 5-aza-dC/TSA treatment resulted in synergistic activation of KLK6 . We show that KLK6 inactivation is associated with hypermethylation of specific CpG dinucleotides located in the KLK6 proximal promoter and overexpression with complete demethylation. These results indicate a causal role of DNA methylation and chromatin structure in cancer-associated loss of KLK6 expression. In some breast cancer cell lines, KLK6 expression could be restored by the vitamin D3 analog EB1089. Our data indicate that transcriptional deregulation of KLK6 in cancer cells during breast cancer progression is complex and certainly not uniform in different tumors, involving epigenetic mechanisms as well as pathways regulated by nuclear receptors. This allows for the pharmacological modulation of KLK6 with potential therapeutic implications.

Only one transcript for KLK10 was identified by RT-PCR in lung tissue, whereas KLK11 expressed at least four alternative transcripts. Quantitative analysis of KLK10 and KLK11 expression levels was assessed by real-time PCR, in a cohort of 47 patients with non-small-cell lung cancer (NSCLC). Expression levels of these genes were widely distributed in the population studied. Multivariate analysis revealed a correlation between KLK10 over-expression and the squamous cell carcinoma histotype ( p =0.034). There was no correlation between gene expression and patient survival. Overall, both genes behaved similarly ( p <0.001). These results suggest a co-regulation of KLK10 and KLK11 expression in lung and a lack of KLK10 suppressor role in NSCLC. Finally, our findings indicate that these genes are likely involved in normal physiology processes in bronchus.

Kallikrein 11 ( KLK11 , TLSP , hippostatin) is a member of the human kallikrein gene family, which includes PSA, KLK2 and 12 other members, all localized on chromosome 19q13.4. The aim of this study was to investigate whether KLK11 expression could be used to discriminate prostate cancer (CaP) from benign prostatic hyperplasia (BPH) in needle prostate biopsies. We analyzed the expression of the prostate-type variant of the KLK11 gene in 64 CaP and BPH tissues obtained by transrectal ultrasound-guided needle biopsy. Reverse transcription (RT), PCR and image analysis methodologies were developed and used. Of the 42 BPH tissues examined, only 10 (23.8%) were positive for prostate-type KLK11 expression, while of the 22 CaP patients, 12 (54.5%) were KLK11 -positive ( p =0.025). Logistic regression and receiver operating characteristic curve analyses demonstrated that KLK11 expression has a significant discriminatory value (crude odds ratio=3.84, p =0.016; area under the curve, 0.65, 95% CI 0.51–0.80) between CaP and BPH in needle prostate biopsies.

The tissue kallikrein gene family consists of 15 genes tandemly arranged on human chromosome 19q13.4. Most kallikrein genes are characterized by aberrant expression patterns in various human cancers, a feature that makes them ideal cancer biomarkers. In the present study, we investigated the effect of the epigenetic drug compound 5-aza-2′-deoxycytidine on the expression of downregulated kallikrein genes in prostate, breast, and ovarian cancer cell lines. Reactivation of multiple kallikrein genes was observed, although some of these genes do not contain CpG islands in their genomic sequence. Epigenetic regulation provides a new mechanism for the pharmacological modulation of kallikreins in human cancers with putative therapeutic implications.

Human kallikrein 11 (hK11) was evaluated in a percentage free PSA-based artificial neural network (ANN) to reduce unnecessary prostate biopsies. Serum samples from 357 patients with (n=132) and without (n=225) prostate cancer (PCa) were analyzed and ANN models were constructed and compared to all parameters. The discriminatory power of hK11 was lower than that of PSA, but receiver operator characteristic (ROC) analyses demonstrated significantly larger areas under the curves for the ANN compared to all other parameters. ANNs with hK11 may lead to a further reduction in unnecessary prostate biopsies, especially when analyzing patients with less than 15% free PSA.

The human tissue kallikrein family of serine proteases (hK1–hK15 encoded by the genes KLK1 – KLK15 ) is involved in several cancer-related processes. Accumulating evidence suggests that certain tissue kallikreins are part of an enzymatic cascade pathway that is activated in ovarian cancer and other malignant diseases. In the present study, OV-MZ-6 ovarian cancer cells were stably co-transfected with plasmids expressing hK4, hK5, hK6, and hK7. These cells displayed similar proliferative capacity as the vector-transfected control cells (which do not express any of the four tissue kallikreins), but showed significantly increased invasive behavior in an in vitro Matrigel invasion assay ( p <0.01; Mann-Whitney U-test). For in vivo analysis, the cancer cells were inoculated into the peritoneum of nude mice. Simultaneous expression of hK4, hK5, hK6, and hK7 resulted in a remarkable 92% mean increase in tumor burden compared to the vector-control cell line. Five out of 14 mice in the ‘tissue kallikrein overexpressing’ group displayed a tumor/situs ratio greater than 0.198, while this weight limit was not exceeded at all in the vector control group consisting of 13 mice ( p =0.017; χ 2 test). Our results strongly support the view that tumor-associated overexpression of tissue kallikreins contributes to ovarian cancer progression.

Accumulated evidence has shown that human tissue kallikreins (hKs), a group of 15 homologous secreted serine proteases, are novel cancer biomarkers. We report here the inhibition profiles of selected hKs, including hK5, hK7, hK8, hK11, hK12, hK13, and hK14, by several common serine protease inhibitors (serpins) found in plasma. The association constants for the binding of serpins to kallikreins were determined and compared. Protein C inhibitor was found to be the fastest-binding serpin for most of these hKs. α 2 -Antiplasmin, α 1 -antichymotrypsin, and α 1 -antitrypsin also showed rapid inhibition of certain hKs. Kallistatin exhibited fast inhibition only with hK7. Our data demonstrate that these hKs are specifically regulated by certain serpins and their distinct inhibition profiles will be valuable aids in various aspects of kallikrein research.

We tested the hypothesis that human tissue kallikreins (hKs) may regulate signal transduction by cleaving and activating proteinase-activated receptors (PARs). We found that hK5, 6 and 14 cleaved PAR N-terminal peptide sequences representing the cleavage/activation motifs of human PAR 1 and PAR 2 to yield receptor-activating peptides. hK5, 6 and 14 activated calcium signalling in rat PAR 2 -expressing (but not background) KNRK cells. Calcium signalling in HEK cells co-expressing human PAR 1 and PAR 2 was also triggered by hK14 (via PAR 1 and PAR 2 ) and hK6 (via PAR 2 ). In isolated rat platelets that do not express PAR 1 , but signal via PAR 4 , hK14 also activated PAR-dependent calcium signalling responses and triggered aggregation. The aggregation response elicited by hK14 was in contrast to the lack of aggregation triggered by hK5 and 6. hK14 also caused vasorelaxation in a phenylephrine-preconstricted rat aorta ring assay and triggered oedema in an in vivo model of murine paw inflammation. We propose that, like thrombin and trypsin, the kallikreins must now be considered as important ‘hormonal’ regulators of tissue function, very likely acting in part via PARs.