Processing-independent analysis (PIA)--a new diagnostic tool

Processing-independent analysis (PIA): a method for quantitation of the total peptide-gene expression

The translational product of protein-coding genes undergoes extensive posttranslational modifications. The modifications ensure an increased molecular and functional diversity at protein- and peptide-level. Prohormones are small pro-proteins that are expressed in many cell types, for instance endocrine cells, immune cells, myocytes and neurons. Here they mature to bioactive peptides (cytokines, hormones, growth factors, and neurotransmitters) that are released from the cells in an often regulated manner. The posttranslational processing of prohormones is cell-specific, however, and may vary during evolution and disease. Therefore, it is often inadequate to measure just a single peptide fragment as marker of endocrine, immune, and neuronal functions. In order to meet this challenge, we developed years back a simple "processing-independent analysis" (PIA) for accurate quantification of the total pro-protein product - irrespective of the degree and nature of the posttranslational processing. This review provides an overview of the PIA principle and describes examples of PIA results in different peptide systems.

Cholecystokinin expression in tumors: biogenetic and diagnostic implications

Cholecystokinin (CCK) is a classic gut hormone. CCK is also a complex system of peptides expressed in several molecular forms in enteroendocrine I cells, in cerebral and peripheral neurons, in cardiac myocytes and spermatozoa. CCK gene expression has now been found at protein or peptide level in different neuroendocrine tumors; cerebral gliomas and astrocytomas and specific pediatric tumors. Tumor hypersecretion of CCK was recently reported in a patient with a metastatic islet cell tumor and hypercholecystokininemia resulting in a novel tumor syndrome, the cholecystokininoma syndrome. This review presents an overview of the cell-specific biogenesis of CCK peptides, and a description of the CCK expression in tumors and of the cholecystokininoma syndrome. Finally, assays for the diagnosis of CCK-producing tumors are reviewed.

Posttranslational processing of progastrin

Gastrin and cholecystokinin (CCK) are homologous hormones with important functions in the brain and the gut. Gastrin is the main regulator of gastric acid secretion and gastric mucosal growth, whereas cholecystokinin regulates gall bladder emptying, pancreatic enzyme secretion and besides acts as a major neurotransmitter in the central and peripheral nervous systems. The tissue-specific expression of the hormones is regulated at the transcriptional level, but the posttranslational phase is also decisive and is highly complex in order to ensure accurate maturation of the prohormones in a cell specific manner. Despite the structural similarities of gastrin and CCK, there are decisive differences in the posttranslational processing and secretion schemes, suggesting that specific features in the processing may have evolved to serve specific purposes. For instance, CCK peptides circulate in low picomolar concentrations, whereas the cellular expression of gastrin is expressed at higher levels, and accordingly gastrin circulates in 10-20-fold higher concentrations. Both common cancers and the less frequent neuroendocrine tumors express the gastrin gene and prohormone. But the posttranslational processing progastrin is often greatly disturbed in neoplastic cells.The posttranslational phase of the biogenesis of gastrin and the various progastrin products in gastrin gene-expressing tissues is now reviewed here. In addition, the individual contributions of the processing enzymes are discussed, as are structural features of progastrin that are involved in the precursor activation process. Thus, the review describes how the processing depends on the cell-specific expression of the processing enzymes and kinetics in the secretory pathway.

The art of measuring gastrin in plasma: a dwindling diagnostic discipline?

The gastrointestinal hormone gastrin is measured in plasma in physiological, pathophysiological and diagnostic investigations. In the diagnosis of hypergastrinaemic diseases such as gastrinomas and gastric achlorhydria, measurement of gastrin concentrations in circulation is crucial. Gastrin circulates, however, not as a single peptide but as a mixture of peptides of different lengths and amino acid derivatizations. Moreover, in hypergastrinaemia the peptide pattern changes. Consequently, diagnostic gastrin measurements require immunoassays that recognize the pathological plasma patterns, which are characterized by a predominance of the large peptides (gastrin-34 and gastrin-71) and less, if any, of the shorter main form of gastrin in normal tissue, gastrin-17. Alternatively, and in specific cases, "processing-independent assays" (PIA) for progastrin may be considered, since hypersecreting gastrin cells also release substantial amounts of biosynthetic precursors and processing intermediates. Recently, gastrin kits that do not take the pathological plasma patterns into account have been marketed and may miss the diagnosis. Therefore, proper diagnosis of gastrinomas and other hypergastrinaemic diseases requires insight into cellular gastrin synthesis and peripheral metabolism, and also into the design of useful immunoassays. This review discusses the art of measuring gastrin in plasma with adequate diagnostic specificity.

Increased gastrin precursor secretion in preterm infants and their mothers at birth

AIM OF THE STUDY

To analyze gastrin and its precursors in plasma from mothers and preterm infants just after birth.

PATIENTS AND METHODS

38 preterm infants, median (range) gestational age 32 wks (25-36 wks) and birth weight 1995 g (538-2764 g), and 29 mothers participated. Thirteen mothers were delivered by Cesarean section. The concentration of the total progastrin product was measured by a new processing-independent analysis (PIA) for progastrin, and the concentration of mature (i.e. alpha-amidated) gastrin was measured by conventional RIA.

RESULTS

The median concentration of the progastrin product was significantly higher than that of alpha-amidated gastrin in both mothers and infants (mothers: 40 versus 8 pmol/l; infants: 67 versus 9 pmol/l). The peptide concentrations were not correlated to sex, birth weight, umbilical cord pH, blood glucose concentrations or placenta weight. In mothers, the progastrin product was higher after multiple births than after a single birth.

CONCLUSION

As more than 90% of the total progastrin product in healthy nonpregnant adults is alpha-amidated, the results indicate that the biosynthesis of progastrin and its products is increased in mothers and even further in infants. Post-translational maturation is, however, markedly attenuated. Since it was recently suggested that progastrin and some of its non-amidated processing intermediates may act as growth factors for the gastrointestinal mucosa, the results of this study indicate that the gastrointestinal mucosa of pregnant women and infants is subject to gross gastrinergic growth stimulation.

Processing-independent analysis in the diagnosis of gastrinomas

BACKGROUND

This study evaluates whether a new analytic principle, processing-independent analysis (PIA), offers better specificity and sensitivity than the conventional gastrin radioimmunoassay in the diagnosis of gastrinomas.

METHODS

Plasma concentrations of alpha-amidated gastrins and the total progastrin product were measured with radioimmunoassay and with PIA, respectively, in 512 samples taken for gastrin measurement and in a selected group of gastrinoma patients (n=10).

RESULTS

Among the 512 patients were 9 with gastrinomas. In plasma from these patients the median degree of amidation (ratio of alpha-amidated gastrins to total progastrin product) was 75% (range, 25-98%), whereas in the other groups the medians varied from 41% to 86%. In the second group of gastrinoma patients all had a degree of amidation of less than 50%.

CONCLUSIONS

In screening for gastrinomas PIA offered no diagnostic advantages in comparison with conventional gastrin radioimmunoassay. However, in selected patients who in spite of normal or slightly increased concentrations of amidated gastrins were still suspected of having gastrinoma, additional measurement of the total progastrin product showed incomplete processing of progastrin and thus proved helpful in establishing the diagnosis.

Gastroenteropancreatic tumours and prohormones

The structures and post-translational maturation of pancreatic and gastrointestinal prohormones are reviewed with emphasis on Danish contributions to today's knowledge. The review describes general, cell-specific, and tumour-specific prohormone-processing patterns. Since prohormone-processing in endocrine tumours is often attenuated, conventional assays that measure only the phenotypic endpoint of hormone gene expression (i.e. the bioactive hormone) do not quantitate tumour activity accurately. In contrast, measurements that include also prohormones and processing intermediates provide more accurate data on hormone synthesis in gastroenteropancreatic endocrine tumours. In order to comply with such demands we have developed a new analytical principle (processing-independent analysis (PIA)) which quantitates the entire translation product irrespective of the degree of processing. The significance of PIA in routine diagnostics awaits prospective evaluation. We hope that the present review illustrates how the tumour biology of endocrine cells in the pancreas and the gut has been an essential research area in Danish gastroenterology and endocrinology--one purpose being improvement of early diagnosis of endocrine tumours in the gut and the pancreas.

Quantitation of procholecystokinin and its products in plasma by processing-independent analysis

A procedure for processing-independent quantitation of procholecystokinin (proCCK) and its products has been applied to plasma. The procedure is based on tryptic cleavage after Lys61 and Arg71 with subsequent monospecific radioimmuno-analysis of fragment 62-71 of human proCCK, which again corresponds to fragment 1-10 of CCK-22. The detection limit of the analysis was 0.2 pmol/l. Plasma was extracted with ethanol. In plasma from 13 healthy volunteers the basal concentration with the above-mentioned radioimmunoassay was 1.1 +/- 0.1 pmol/l (mean +/- S.E.M.) before, and 13.7 +/- 0.6 pmol/l after, incubation with trypsin. Two hours after ingestion of a mixed meal, the plasma concentration was 2.0 +/- 0.1 pmol/l before, and 21.7 +/- 1.2 pmol/l after tryptic cleavage. With a conventional CCK radioimmunoassay specific for the C-terminally amidated and O-sulfated bioactive epitope, the concentration was 1.0 +/- 0.1 pmol/l in the basal state and 4.2 +/- 0.4 pmol/l 2 h after a meal. Tryptic cleavage did not increase the concentrations of amidated, bioactive CCK peptides. In plasma from 37 patients with the carcinoid syndrome, the basal concentration of proCCK and its products was 14.1 (2.8-150.4) pmol/l (median (range)), compared with 0.3 (0-18.8) pmol/l for carboxyamidated CCK. Only two patients had significantly elevated CCK concentrations. We conclude that processing-independent analysis is useful for quantitation of proCCK and its products in plasma, since it quantitates CCK cell secretion more accurately than conventional CCK assays.

A processing-independent assay for human procholecystokinin and its products

In order to develop a processing-independent analysis for procholecystokinin (proCCK) and its products, antibodies were raised against the synthetic fragment 62-71 of human proCCK. All rabbits (n = 8) responded to the immunization. One (No. 89,009) produced antibodies of particularly high titer (1:350,000), homogeneity (Sips' index approximately 1.0) and binding affinity (K0 eff approximately 0.88 x 10(12) l/mol). A radioimmunoassay using this antiserum and [125I]tyrosine-extended fragment 62-71 measured the total CCK mRNA product after tryptic cleavage at Lys61 in normal and neoplastic tissue independent of the degree of precursor processing. In addition to previously known CCK producing tumors, CCK was found also in a thoracic round-cell tumor (Askin tumor) and in brain tumors (gliomas and astrocytomas). These tumors processed proCCK poorly. Thus, they contained 11 and 23 (mean n = 5) pmol/g of proCCK and its products before, versus 71 and 99 (mean) pmol/g after tryptic cleavage, respectively. Accordingly, gel chromatography revealed significant amounts of unprocessed proCCK, large molecular forms of glycine-extended CCKs and the well-known carboxyamidated and tyrosine O-sulfated bioactive CCK-83, -58, -33, -22 and -8. We conclude that monospecific antibodies directed against the N-terminus of sequence 62-71 of human proCCK are suitable for processing-independent analysis (PIA) for proCCK and its products. Moreover, we suggest that such PIA should be used for quantitation of CCK gene expression at peptide level in normal tissue and tumors.

Measurement of gut hormone gene expression: mRNA and peptides

During the past decade numerous methods for measurement of mRNA and peptides have been developed. Since the expression cascade from DNA to protein is regulated at all levels, the methods should be carefully designed to accomplish the purpose of the analysis. Regulation of the nuclear processing, the translational activity and the decay of a particular mRNA changes the proportionality between transcriptional activity and production of prepropeptide. Moreover, the post-translational maturation of the pro-hormones may be attenuated. Detection of mRNA is valuable and feasible because it is easy to generate cDNA probes for most hormones, and because mRNA demonstration unequivocally indicates the cellular site of gene expression. The deduction of preprohormone structures has also made it possible to improve the versatility of radioimmunoassays (RIA). Monospecific antibodies and pure tracers have allowed the development of sequence-specific RIA libraries for bioactive peptides and their precursors. Recently we have introduced a simple processing-independent analysis (PIA) for clinical use, since the post-translational maturation of gut peptides may be changed in gastrointestinal diseases. So far PIA has improved the diagnostic sensitivity for gut hormone tumours.

Helicobacter pylori infection and exaggerated gastrin release. Effects of inflammation and progastrin processing

Helicobacter pylori infection is associated with exaggerated gastrin release. We investigated whether this abnormality was due to the bacteria or the immune response. Fasting and meal-stimulated 'total' and amidated gastrin were measured in 10 H. pylori-infected volunteers before eradication therapy, after 2 and 14 days of therapy, and 4 weeks after completion of therapy. The exaggerated meal-stimulated gastrin concentration remained unchanged after 2 days of therapy, although the polymorphonuclear cell infiltrate and H. pylori bacteria were no longer evident. The expected fall in gastrin concentration after 14 days of therapy was associated with a reduction in the density of mucosal mononuclear cells, suggesting exaggerated gastrin release was related to chronic inflammation or to H. pylori or its products. The effect of H. pylori on normal progastrin processing was also assessed; 2 control groups were included: 10 H. pylori-uninfected volunteers and 13 patients with H. pylori peptic ulcers. There was a significant difference in the proportion of circulating gastrins that were biologically active amidated gastrins between ulcer patients and uninfected controls (56.7 +/- 4% versus 33.8 +/- 4%, p < 0.001). The proportion of amidated to total gastrins did not increase after successful eradication.

Expression but incomplete maturation of progastrin in colorectal carcinomas

BACKGROUND

To evaluate the hypothesis that gastrin is a local growth factor in colonic carcinomas, the expression of gastrin messenger RNA (mRNA) and peptides were examined in five human colon carcinoma cell lines, 12 solid colon carcinomas, and normal colonic tissue.

METHODS

Northern analysis, reverse-transcription PCR, and a library of sequence-specific radioimmunoassays were the principal methods.

RESULTS

Cell lines, tumors, and normal tissue all expressed a gastrin mRNA of 0.7 kilobases, and all cell lines contained incompletely processed progastrin (range, 17-54 fmol/10(6) cells). Two cell lines secreted progastrin into the media (LoVo, 25 +/- 3 pmol/L; HCT116; 12 +/- 2 pmol/L). Normal colonic tissue and all the solid tumors also contained progastrin, the concentration being higher in tumors (range, 0.4-2 pmol/g) than in normal tissue (range, 0.1-0.2 pmol/g). Only one tumor contained carboxyamidated gastrins.

CONCLUSIONS

Normal and neoplastic colonic mucosa both express the gastrin gene, but the posttranslational phase of expression is attenuated. The incomplete processing and low level of expression suggest that autocrine gastrin secretion has only minor significance for normal adult and most neoplastic colonic tissue.

Co-transcription of the gastrin and cholecystokinin genes with selective translation of gastrin mRNA in a human gastric carcinoma cell line

So far, no cells have been found to synthesize both of the homologous hormones, cholecystokinin and gastrin. Northern analysis and reverse transcription PCR showed, that the human gastric carcinoma cell line (AGS) expresses both a gastrin mRNA of 0.7 kb and a cholecystokinin transcript of 0.8 kb. A library of sequence-specific radioimmunoassays, cleavage with processing-like enzymes and chromatography subsequently revealed that the gastrin mRNA was translated into progastrin that was constitutively secreted into the medium (45 +/- 3 pmol/l). Neither procholecystokinin nor any of its processing products were detectable in cells and media. The results suggest that differentiation into gastrin- or cholecystokinin-producing cells may be regulated at the translational level. The gastric cell line, AGS, provides a model for studies of translational regulation of cell differentiation.