Cholecystokinin: Clinical aspects of the new biology

Abstract

Cholecystokinin (CCK) is a classic gut hormone that has been known for almost a century to regulate gallbladder emptying, pancreatic enzyme secretion, and gastrointestinal motor activity. In 1968, the CCK structure was identified by Viktor Mutt and Erik Jorpes from porcine gut extracts as a peptide of 33 amino acid residues. Based on that structure, physiological, immunochemical, molecular, and cell biological research has since expanded the insight into the biology of CCK remarkably. Thus, CCK was the first identified intestinal satiety signal to the brain. Moreover, the CCK gene is now known to be expressed in different molecular forms not only in the gut, but very much so in central and peripheral neurons, in addition to extra‐intestinal endocrine cells, immune cells, cardiomyocytes, spermatogenic cells, and certain fat cells. Accordingly, CCK peptides function not only as hormones. They are also neurotransmitters, paracrine growth and satiation factors, anti‐inflammatory cytokines, incretins, adipokins, myokines, potential fertility factors, and tumor markers. Consequently, CCK biology has now opened windows for insights into pathophysiology with diagnostic and therapeutic possibilities in metabolic disorders (obesity, eating disorders, and diabetes mellitus), gallbladder disease, neuropsychiatric diseases (cerebral tumors, memory, and anxiety disorders), cardiac diseases (prognosis in heart failure), neuroendocrine and pediatric tumors, as well as perhaps infertility.

Introduction

Endocrinology covers a wide range of different hormones: steroids, monoamines, proteins, and peptides. In modern peptide endocrinology, gastrointestinal hormones have recently surfaced as a strikingly dynamic area with considerable clinical impact [1]. It suffices to mention, for example, glucagon‐like peptide‐1 (GLP‐1) and glucose‐dependent insulinotropic polypeptide (GIP), from which several drugs have already been derived and used widely in the treatment of diabetes and obesity, with additional candidates in the pipeline and potential for therapy of cardiovascular, renal, and cerebral diseases as well [2, 3, 4].

GIP and GLP‐1 are well described as polypeptides secreted from intestinal K‐ and L‐cells, respectively [1]. Other gastrointestinal hormones, however, have a more complex biology. Among these is not least cholecystokinin (CCK), whose role as a hormone in digestion may turn out to be secondary compared to some of the extraintestinal activities implicated in the new biology.

Following a summary of the CCK history, this article will offer a brief overview of the new CCK biology, including phylogeny, biogenesis, cellular expression with cell‐specific peptide patterns, receptors, and subsequently derived functions. A note on reliability of CCK measurements will also be given before describing the perspective of pathophysiological involvement of CCK in metabolic disorders, gallbladder pathology, neuropsychiatric diseases, cardiac diseases, and various neoplasias.

A short history

More than a century ago, some European physiology laboratories observed that administration of acid into the canine duodenum stimulated bile secretion from the liver and gallbladder contraction [5, 6, 7, 8; for review, see ref. 9]. For decades, the question was whether the effect was due to the first discovered gut hormone, secretin [10], or to a yet unknown hormone. In 1928, however, Ivy and Oldberg provided evidence that the gallbladder effect could not be ascribed to secretin. Consequently, implication of a new gut hormone was necessary to explain the activity. Ivy and Oldberg named the hormone CCK [11]. Subsequently, a few laboratories tried to purify CCK from the small intestine—in vain, however, because the necessary biochemical purification technologies were not yet available. Therefore, not until Viktor Mutt and Erik Jorpes in Stockholm in the 1950s and 1960s established a large plant for extraction and purification from vast amounts of porcine jejunal mucosa (20 km small intestine) did structure identification become possible. Hence, Mutt and Jorpes identified a peptide with CCK activity. The structure was published in 1968 [12]. It had a sequence of 33 amino acid residues, a carboxyamidated C‐terminal sequence homologous to that of gastrin, and besides, it was tyrosyl O‐sulfated (Fig. 1). Elucidating the structure of this peptide became a decisive milestone. With in vitro synthesized fragments of CCK‐33, plenty of material became available for physiological studies, for production of antibodies to be used in immunoassays and immunohistochemistry, as well as for identification of the CCK gene and new molecular forms of CCK peptides. These tools together with other methods from molecular and cell biology paved the way for a number of paradigmatic shifts that have unfolded a surprisingly multifaceted biology of CCK. Consequently, CCK is now seen as a ubiquitous peptide messenger system in the body [13]. The basic chemistry and biology has recently been detailed elsewhere [14]. Below follows a summary of these aspects as premises for the discussion of clinical perspectives. First, however, two useful definitions are presented.

Fig. 1.

The legacy of Viktor Mutt, that is, the structure of the 33 amino acid residues of porcine cholecystokinin as published by Mutt and Jorpes in 1968 (see ref. [12]). Note that the tyrosyl residue in position 27 is O‐sulfated and that the phenylalanyl residue in position 33 is carboxyamidated. The C‐terminal boxed sequence is the bioactive, receptor‐bound epitope of CCK.

Definitions

The CCK system

The system comprises the specific molecular elements in the biogenesis and function of CCK. Thus, gene, mRNA, biosynthetic precursors (prepro‐ and proCCK), processing intermediates, and mature bioactive peptides (CCK‐83, ‐58, ‐33, ‐22, ‐8, and ‐5), as well as receptors (CCK₁, CCK₂, and CCK₃), constitute the CCK system.

The biology of CCK

At first, the biology was simple: CCK was a single peptide (CCK‐33) released from a specific endocrine gut cell to regulate gallbladder contraction, somewhat in the same way as insulin and glucagon are singular peptides released from specific pancreatic islet cells to regulate metabolism. The new biology, however, is the sum of additional evolutionary, biochemical, and physiological insights obtained since Mutt and Jorpes identified CCK‐33 in gut extracts [12]. Several of the new insights—for instance, cerebral neurotransmitter effects, gut‐brain satiety effects, and anti‐inflammatory cytokine effects—may almost overshadow the significance of the original concept of CCK as just a regulator of gallbladder contraction [14].

The new biology of CCK

Phylogenesis

CCK is a member of a family that in mammals includes the antral hormone gastrin [15]; in amphibians, the skin peptide caerulein [16, 17]; and in protochordates, the neuropeptide cionin [18]. Insect neuropeptides, the sulfokinins [19, 20], are also structurally related and therefore included in the family (Fig. 2). Evolutionary studies of the genes indicate that CCK peptides originated almost 600 million years ago [21, 22, 23, 24].

Fig. 2.

The decisive homologous structure in members of the cholecystokinin peptide family in mammals, amphibians, protochordates (Ciona intestinalis), and insects (a cockroach (Leucophaea madera) and a fruit fly (Drosophila melanogaster). Note that the boxed C‐terminal tetrapeptide sequences are all carboxyamidated and that this tetrapeptide amide as such is the core structure necessary for receptor binding. Note also O‐sulfation of the more N‐terminal tyrosyl residues. The tyrosyl‐sulfation is a derivatization that modifies the affinity and hence the specificity of receptor bindings.

Biogenesis and molecular heterogeneity of peptides

There is one CCK gene [25, 26], which is transcribed to a single mRNA of 750 bases [25, 26, 27]. The mRNA molecule is then translated to a precursor protein (preproCCK) with a sequence of 115 amino acid residues [27]. The N‐terminal signal or pre‐sequence is probably removed during translation, leaving a proCCK of 95 amino acids. ProCCK subsequently undergoes extensive posttranslational modifications by a multitude of processing enzymes acting along cellular secretory or synaptic pathways [28, 29, 30, 31, 32, 33, 34, 35, 36, 37]. The covalent modifications comprise endoproteolytic cleavages at mono‐ and dibasic sites, amino acid derivatizations, and N‐ or C‐terminal trimmings. Notably, none of the modifications are complete. As a result of the posttranslational processing, a number of different bioactive CCK peptides are accumulated in the secretory granules (endocrine cells) or presynaptic vesicles (neurons): CCK‐83, ‐58, ‐33, ‐22, ‐8, and ‐5. The first five occur in both O‐sulfated and nonsulfated forms, whereas CCK‐5 (without a tyrosyl residue) is unsulfated. The peptide patterns, however, are cell‐specific (for recent reviews, see refs. [14, 37]).

The widespread tissue and cell expression

In accordance with the classical role as gut hormone, specific endocrine cells in the intestinal mucosa, named I‐cells, have been identified in the duodenum and proximal jejunum as the major source of intestinal CCK peptides [38, 39, 40]. Presumably, a majority of CCK in plasma originates from these cells. The I‐cells, however, are present also in the distal jejunum, in the entire ileum, and sporadically in the mucosa of the proximal colon [41, 42]. Because the duodenum is very short in comparison with the remaining small intestine, jejunum as well as ileum each produce considerably more CCK than the duodenum [41, 42, 43 and unpublished results]. Thus, it is misleading to characterize CCK primarily as a duodenal hormone, as still seen in several textbooks.

Most CCK in mammals is, however, not synthesized in the gut but in neurons, central as well as peripheral [30, 37, 44, 45, 46, 47, 48]. Estimated by concentrations of CCK mRNA, the brain expresses four–five times more CCK than the small intestinal mucosa (unpublished results). The cerebral CCK neurons also express “classic” transmitters, such as GABA in interneurons and glutamate in pyramidal and thalamic neurons [48]. Moreover, the gut and pancreatic islets contain CCK neurons, especially abundant in the colon [44, 49].

Although cerebral neurons and intestinal I‐cells are clearly the two major expression sites, specific expression has in addition been detected in other cells (Table 1). They include pituitary corticotrophs [50, 51], thyroid C‐cells [52, 53], adrenal medullary cells [54], spermatogenic cells [55, 56], cardiac myocytes [57], circulating monocytes [58, 59, 60], and fat cells, at least in some mammalian species (unpublished results).

Table 1.

Tissue concentrations of bioactive cholecystokinin (CCK) peptides in mammals.

Tissue	Tissue content a (pmol/g)
Intestinal Tract:
Duodenal mucosa	200
Jejunal mucosa	150
Ileal mucosa	20
Colonic mucosa	5
Central Nervous System:
Cerebral cortex	400
Hippocampus	350
Hypothalamus	200
Cerebellum	2
Spinal cord	40
Peripheral Nervous System:
Vagal nerve	25
Sciatic nerve	15
Nerveplexes in colonic wall	5
Extraintestinal Endocrine Glands:
Adenohypophysis	25
Neurohypophysis	20
Thyroid gland	2
Adrenal medulla	1
Urogenital Tract:
Testicles	5
Spermatozoas	1
Cardiovascular System:
Atrial myocytes	10
Ventricular myocytes	2
Mononuclear Immune Cells b :	++

^a Orders of magnitude based on measurement of tissue extracts from several mammalian species.

^b Expression determined by immunocytochemistry of monocytes.

Cell‐specificity of CCK peptide expression

As shown in Table 2, there are at least three different molecular patterns of bioactive peptides in cells expressing the CCK gene. First, in both central and peripheral neurons, O‐sulfated CCK‐8 and the unsulfated CCK‐5 predominate. A similar pattern is seen in adrenal medullary cells (which are neuronally derived). Second, gut endocrine I‐cells express a considerably broader pattern that includes large molecular forms (CCK‐58, ‐33, and ‐22) as well as short forms (CCK‐8 and ‐5) [61], where CCK‐33 is the predominant form in humans [62], but CCK‐58 in dogs [63]. The four longest CCK peptides in I‐cells all occur in sulfated and nonsulfated variants [35]. Finally, unique and specific patterns are found in lower concentrations in pituitary corticotrophs, thyroid C‐cells, spermatogenic cells, cardiac myocytes, and circulating monocytes [50, 51, 52, 53, 54, 55, 56, 57, 58, 59].

Table 2.

Cell‐specific patterns in mammals of predominant cholecystokinin (CCK) peptides.

Cells	Peptide pattern
Intestinal endocrine cells (I‐cells)	CCK‐58, ‐33, ‐22, ‐8, and ‐5 (s and ns)
Cerebral neurons	CCK‐8 (s) and CCK‐5
Peripheral neurons	CCK‐8 (s) and CCK‐5
Pituitary corticotrophs	CCK‐83 (s), CCK‐58 (s), and CCK‐33 (s)
Thyroid C‐cells	CCK‐8 (ns) and CCK‐5
Adrenal medullary cells	CCK‐8 (ns and s)
Spermatozoas	CCK‐22 (ns and s) and CCK‐8 (ns)
Cardiomyocytes	proCCK (25‐94) and CCK‐8 (s)

CCK receptors

Using isotope‐labeled CCK fragments or analogues as agonists as well as synthetic antagonists, the distribution and distinction between “alimentary” (CCK_A) and “brain” (CCK_B) receptors were studied already in the 1970s and 1980s [64, 65]. In 1992, however, the structures of both the CCK_A and the gastrin/CCK_B receptors were deduced after cDNA cloning [66, 67]. The receptors are now named CCK₁ and CCK₂, respectively. Later, informative reviews of receptor distribution, physiology, and pharmacology have been published [see refs 68, 69]. Very recently, however, a third CCK receptor (GPR 173) has also been identified in cerebral tissue [70]. Further information about expression of this CCK₃ receptor also in extracerebral tissue remains to be published. The binding specificity of the CCK₁ and CCK₂ receptors differs. Thus, all carboxyamidated CCK and gastrin peptides, irrespective of sulfation, are bound with similar affinities to the gastrin/CCK₂ receptor, whereas the CCK₁ receptor is more selective and binds only tyrosyl O‐sulfated CCK peptides (Fig. 3).

Fig. 3.

The homologous biosynthetic precursors, proCCK and progastrin, are both processed to O‐sulfated (s) as well as non‐sulfated (ns) peptide agonists for the cholecystokinin (CCK) receptors. Only sulfated CCK peptides are bound with high affinity to the CCK₁ receptor, whereas the promiscuous CCK₂ receptor binds all carboxyamidated products of both proCCK and progastrin.

CCK functions

An overview of the multiple functions of the CCK system requires recognition of the new biology in its entirety: On one hand, there are the different cellular origins and release mechanisms for bioactive CCK peptides. On the other hand, there are the targets in terms of CCK receptors and the routes by which the peptides travel to reach the receptors on the target cells (the endocrine, neurocrine, paracrine, autocrine, spermiocrine secretory ways, and synaptic transmissions). The CCK receptors are expressed in a majority of organs and tissues in the body: central and peripheral neurons as well as glial cells; the entire gastrointestinal tract; myocytes of the gallbladder and the sphincter Oddi muscles; exocrine and endocrine pancreatic cells; the cardiovascular system; the urogenital system; and the immune system [68, 69]. Consequently, CCK peptides have many labels and may act independently as neurotransmitters, hormones, growth factors, satiation factors, cytokines, myokines, and perhaps adipokines and fertility factors.

On top of this picture, it is of fundamental significance to realize that the target cells for CCK peptides also express receptors for a multitude of other bioactive peptide systems. Therefore, different peptide systems interact and cross‐talk along intracellular signal‐transduction pathways. Well‐known examples are the mutual potentiation of CCK peptides with secretin in their stimulation of exocrine pancreatic enzyme and bicarbonate secretion [71, 72]; the CCK/gastrin potentiation of GLP‐1 in stimulation of pancreatic beta‐cell growth [73]; the interaction of CCK with GLP‐1 and PYY in signaling to the brain via afferent vagal fibers to regulate food intake [74]; and, for instance, the balance between sulfated and nonsulfated CCK peptides from CCKomas in the inhibition or stimulation of gastric acid secretion [75] (Fig. 4). It is likely that studies of the interactive physiology between CCK and other bioactive peptide systems, as well as the interaction with non‐peptidergic extracellular messenger molecules, will be an essential part of the future physiology and pharmacology of CCK [76].

Fig. 4.

Cholecystokinin (CCK) influences gastric acid secretion: CCK peptides inhibit acid secretion in normal healthy subjects and animals after binding of blood‐borne sulfated CCK to the CCK₁ receptor on gastric D‐cells. Subsequently, the D‐cell release via somatostatin to the somatostatin₂ receptor on parietal cells blocks further acid secretion. Although both sulfated and non‐sulfated CCK peptides also are bound to the CCK₂ receptor on ECL and parietal cells, their acid stimulatory effect is marginal in comparison with that of gastrin that circulates in 10‐fold higher concentration in healthy subjects. In CCKoma patients (with massive overproduction of non‐sulfated CCK peptides (refs. [176, 177]), the effect is opposite, that is, hypersecretion of gastric acid and development of duodenal ulcers. The acid stimulatory effect in these patients occurs because the CCK₁ receptor cannot bind non‐sulfated CCK peptides, and thus the D‐cell does release only limited amounts of the inhibitory somatostatin. In other words, non‐sulfated CCK peptides act like gastrin peptides.

Measurement of CCK

It is obvious that progress in understanding the basic CCK biology as well as the roles of CCK in pathophysiology, diagnostics, and therapy has been and henceforth will have to be based on reliable measurements. So far, such reliability has been a problem. Therefore, a comment about the premises for accuracy of CCK assays and about the methods required to obtain the desired accuracy:

Assay premises

CCK is a most, if not the most, difficult peptide system to measure accurately in biological fluids, particularly so in plasma. The difficulty is caused by the biochemistry of CCK: first, the molecular heterogeneity with different peptides (CCK‐58, ‐33, ‐22, and ‐8) requires assays that measure these forms with equimolar potency. Second, the concentrations of CCK peptides in circulation are very low for a hormone. In mammals, the concentrations in fasting are often femtomolar. After meals, they increase to a few picomolar levels (3–5 pM). Measurement of these levels require high‐affinity antibodies in the immunoassays. Third, the related sister hormone, gastrin (Fig. 2), circulates in 10‐fold higher concentrations in plasma. Consequently, far most antibodies raised for immunoassays against CCK will cross‐react with gastrin peptides. Finally, the high protein concentrations in plasma may interfere unspecifically with immunoassay measurements. Therefore, preanalytical extraction from protein‐rich fluids is necessary.

Assay methods

There are three relevant methodologies to consider for CCK measurements in biological fluids (blood plasma, cerebrospinal fluid, semen, and tissue extracts). Bioassays have historically been crucial for the initial purifications from intestinal tissue and, for instance, amphibian skin tissue extracts [12, 16, 77]. However, in general, bioassays suffer from insufficient specificity and sensitivity for plasma measurements. One exception has nevertheless been reported [78]. It is based on amylase secretion from isolated pancreatic cells and is sufficiently sensitive and specific for plasma measurements. However, in spite of the impressive reliability, it has been too complex, labor intensive, and costly for practical purposes and hence for wider applicability. An alternative principle of sufficient sensitivity and practicability is the immunoassay technology, especially radioimmunoassay methods (RIA). Establishment of RIAs of sufficient reliability for CCK has, however, as described above, also been a challenge. This is abundantly illustrated by many unspecific and hence misleading commercial CCK immunoassay kits that have been and still are on the market [79, 80]. Nevertheless, a few specific antisera for CCK RIAs have been raised in university laboratories. These RIAs do not cross‐react with gastrin (neither in sulfated nor non‐sulfated forms of gastrin) [79, 81, 82, 83, 84, 85, 86]. Moreover, most of these seem to measure the different molecular forms of bioactive CCK peptides in plasma and tissue extracts adequately [79, 86]. The third assay principle to consider is mass spectrometry (MS). For measurement of bioactive CCK peptides that in the basal conditions circulate in plasma in femtomolar concentrations, MS methods have, however, so far been useless due to insufficient analytical sensitivity. In addition, they are also costly. Nevertheless, an MS‐based method using statistical modeling software was recently reported to be able to measure an N‐terminal fragment of proCCK as a surrogate parameter of CCK secretion into plasma in human subjects [87]. This is possible because proCCK and some of its intermediate processing fragments are known to circulate in substantially higher concentrations than the bioactive CCK peptides [88, 89]. Thus, further improvements in MS technology may be relevant to follow. Additional details in the measurement of CCK is discussed elsewhere [90, 91].

Progress in knowledge about the role of CCK in diseases in general has been slow in comparison with that of most other peptide hormones because of the limitations in the availability of assays that measure the true concentrations of CCK [79, 86]. As a conclusion of the assay discussion, however, the recommendation of today will be to use an entirely specific RIA with an antiserum having characteristics similar to the one detailed in ref. [86].

Pathophysiology of CCK

CCK in obesity and eating disorders

CCK was the first discovered intestinal satiety signal and is by now well‐established as such [92]. It reaches the brain via CCK₁ receptors expressed on afferent vagal fibers [74, 93, 94, 95]. Consequently, there has been an impressive interest in measuring intestinal CCK secretion in human obesity before as well as after various treatments (diets, exercise, gastric bypass, and sleeve surgery) [96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114]. Moreover, intestinal CCK secretion has also been examined in patients suffering from eating disorders (anorexia nervosa, anorexia in the aging, and bulimia nervosa) [115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127].

The obesity studies showed that the plasma CCK concentrations both in the fasting state and after intake of mixed meals are the same in lean and obese people. Neither weight nor sex apparently influences the secretion of CCK [105]. Diet plus exercise‐induced weight loss in obese patients maintained over 2 years may increase CCK secretion both in the basal state and after meals, but only slightly [108]. Weight loss in severely obese patients after gastric bypass operations, however, influences the secretion of CCK. Hence, the basal secretion in the fasting state decreases to concentrations from 1 pmol/L before to 0.35 pmol/L after the operation. After ingestion of mixed meals, however, the peak concentrations increase from 3–5 pmol/L preoperatively to 7–8 pmol/L postoperatively [101, 103, 106, 111]. Hence, in unoperated obese patients, as in lean subjects, there is a five‐fold increase in postprandial CCK concentration. In contrast, the postprandial CCK concentrations increase by 20‐ to 25‐fold in bypass‐operated patients. Such difference renders it likely that meal‐induced CCK secretion contributes to the increased postoperative incretin effect, which in 80% of the bypassed patients reduces or eliminates type 2 diabetes mellitus.

In anorexia nervosa, the CCK results are ambiguous, which to some extent may be explained by assay differences and small numbers of the mainly young girls who have been studied [116, 117, 118, 119, 120]. Generally, there seems to be an increased CCK secretion in the young anorexia patients. The increase, however, is hardly causal but might interfere with refeeding and recovery [128].

Anorexia is quite common among older people. It has been estimated to vary between 15% and 30% and is even higher in nursing homes [121]. Both fasting and food‐stimulated CCK secretion seems to increase with age [122]. Moreover, the effect of exogenous CCK on food intake also appears increased in aging [123], indicating that CCK may play a role, although hardly a major one [128].

CCK secretion in bulimia nervosa has also been examined [124, 125, 126, 127]. Some studies found decreased basal and postprandial CCK concentrations in plasma but also slower gastric emptying. It is therefore possible that the decreased plasma levels to some extent are due to delayed gastric emptying with subsequently reduced food stimulation of the CCK cells in the proximal gut.

In conclusion, the changes in CCK secretion and effects found in obesity and eating disorders are modest and as such hardly decisive. However, it is possible that the significance of even minor variations should be sought in the interaction/potentiation of CCK peptides with other peptide messengers such as GLP‐1 and PYY, as suggested elsewhere [76].

CCK in diabetes mellitus

CCK peptides have evoked some interest in diabetes research, not least because CCK, in itself being a minor incretin, may potentiate the effect of major incretins such as GLP‐1 and GIP. Details of this area are described in recent reviews [129, 130]. Regarding intestinal CCK secretion in diabetes, a reduced postprandial CCK response to low‐fat meals has been observed in patients with type 2 diabetes [131, 132]. The responses to medium‐ and high‐fat meals, however, are normal [132]. Moreover, in accordance with the intestinal mucosal CCK concentrations and I‐cell densities in patients with type 2 diabetes, these parameters are also normal [43]. In type 1 diabetes, abnormal intestinal CCK release has so far not been encountered. Thus, the intestinal CCK synthesis and secretion are apparently not changed to any significant extent in diabetes.

The major premise for associating CCK with diabetes is, nevertheless, the expression of CCK receptors and peptides as well as the sister hormone gastrin in pancreatic islets. Such expression has also been found in experimental diabetes in rodents as well as in human diabetes [73, 133, 134, 135, 136, 137, 138, 139, 140, 141]. On this background, promising anti‐diabetic effects of some modified CCK‐8 analogues (not least [pGlu‐Gln] CCK‐8) have been demonstrated [73, 142, 143, 144, 145]. However, for attempts to treat diabetes in humans with CCK receptor agonists, for safety reasons [146], it has to be low doses of the agonist in combination with analogues of the major incretins such as GLP‐1 and GIP (for review, see ref. [129]).

CCK, gallbladder, and exocrine pancreatic diseases

The association between CCK and gallbladder function is, per definition, the basis for the century‐old biology of CCK, as reflected also in the name of the hormone [11]. Today it is an established fact that CCK peptides from I‐cells in the proximal small intestine in a concentration‐related manner control the intestinal phase of gallbladder emptying via CCK₁ receptors expressed on gallbladder muscles (for recent review, see ref. [147]). The mechanistic details and the relationship to gastrointestinal disorders, pancreatic insufficiency, and gallstone/gallbladder diseases have systematically been studied, not least by Masclee and Lamers et al. [147, 148, 149, 150, 151, 152, 153].

The results are summarized in Table 3, which shows that the occurrence of gallstones frequently accompanies the examined disorders. Although gallstone attacks are painful and annoying, surgical gallbladder removal is today fast and uncomplicated, and life without a gallbladder is pretty normal. Remember that the successful rodent, the rat, does not have a gallbladder. Thus, perhaps the pancreozymic effect of CCK should be considered more essential for digestion than the gallbladder‐emptying effect, as demonstrated by normal rats.

Table 3.

Cholecystokinin secretion and gallbladder motility in some gastrointestinal disorders. ^a

Disorder	Stimulus	CCK secretion	Gallbladder motility	Gallstone prevalence	Mechanism
Partial gastrectomy	Oral meal	Increased	Increased	Increased	Rapid gastric emptying
Sleeve gastrectomy	Oral meal	Increased	Normal	–	Rapid gastric emptying
Roux‐en‐Y bypass	Oral meal	Increased	Decreased	Increased	Rapid gastric emptying with bypass of prox. gut
Celiac disease	Oral meal	Decreased	Decreased	Increased	–
Exocrine pancreatic insufficiency	Oral meal	Decreased	Decreased	Increased	Delayed intestinal fat digestion
Whipple operation	Oral meal	Decreased	Gallbladder removed	–	Reduced pancreatic enzyme secretion
Colectomy	Oral meal	Increased	Normal	Increased	Changes in bile composition
Parenteral nutrition	No enteral feeding	Decreased	Stasis	Increased	No enteral stimulus

^a See also ref. [147] for further details.

CCK in neuropsychiatric diseases

As already mentioned, the majority of CCK peptides in mammals, including humans, are synthesized as neurotransmitters in neurons [30, 31, 32, 47]. Thus, the cerebral regions (except the cerebellum) are loaded with CCK neurons, short interneurons as well as long‐reaching neurons [30, 44, 45, 46, 47, 48]. Considering the hundreds or thousands of contacts each neuron has with other neurons, glial cells, blood vessels, and so on, it looks like an almost insurmountable challenge to identify and rule out the CCK circuits in the brain and how they are involved in the pathophysiology of neuropsychiatric diseases.

Some superficial hints, however, have been given by comparing CCK concentrations in cerebrospinal fluid from patients and healthy controls. Thus, in patients with depression, schizophrenia, Parkinson's disease, and severe alcohol dependence, the cerebrospinal CCK concentrations are decreased but still overlapping with normal levels [154, 155, 156, 157]. In contrast, the CCK levels are increased in multiple sclerosis [158]. Moreover, CCK‐knockout (KO) mice contribute to information about the role of cerebral CCK. Hence, global KO of the CCK gene shows that mice have lost their memory and exhibit increased anxiety [159, 160], in addition to peripheral disturbances of digestive and endocrine functions [159, 160, 161, 162] that also may interfere with memory functions [163, 164]. The increased anxiety in the CCK‐KO mice is paradoxical, because it is now well established that peripheral administration of exogenous CCK‐4 and CCK‐8 in a dose‐related manner induces panic attacks of anxiety in mammals, among which human subjects have been carefully studied [165, 166, 167]. A further player in cerebral anxiety mechanisms is obviously the CCK₂ receptor. Thus, KO mice without this receptor show less anxious behavior in comparison with wild‐type mice [168]. On top of the role of CCK in anxiety and panic disorders, it has now been suggested that CCK peptides may be promising drug candidates for the therapy of Alzheimer's and Parkinson's diseases [169]. Finally, it also deserves mention that CCK in dopaminergic mesolimbic neurons projecting to the frontal brain might be of relevance for the pathophysiology of schizophrenia [170].

CCK in tumors

Like other gut hormones, bioactive CCK peptides are also growth factors [171, 172]. Furthermore, because CCK to a variable extent is expressed in neuroendocrine tumors and may function as tumor marker, a short presentation of the oncogenetic occurrence of CCK may have clinical interest (for reviews, see refs. [173, 174]). Of new and particular clinical interest is the rare pancreatic islet‐cell carcinoma with massive hyperCCKemia that causes a specific CCKoma syndrome [75, 175, 176, 177]. The major symptoms of the syndrome are severe non‐watery diarrhea, weight loss, recurrent peptic ulcers, and gallbladder disease with repeated gallstone attacks [176, 177]. The duodenal ulcer disease and diarrhea in spite of permanently low gastrin concentrations in plasma suggest that some CCK peptides from the pancreatic CCKoma and its metastases may induce gastrinoma‐like symptoms, that is, Zollinger–Ellison syndrome [75]. However, sulfated CCK peptides inhibit gastric acid secretion via CCK₁ receptors on fundic somatostatin cells (Fig. 4). In contrast, non‐sulfated CCK peptides stimulate gastric acid secretion via CCK₂‐receptors on gastric ECL cells. In other words, non‐sulfated CCK acts like gastrin. CCKoma cells may release manyfold more nonsulfated than sulfated CCK peptides [75, 177]. In addition, CCK is also expressed at lower levels in pituitary Cushing and Nelson tumors [178], in thyroid C‐cell carcinomas [53], and in pheochromocytomas [54]. Moreover, increased proCCK expression has been encountered in pediatric round cell Askin tumors [89], neuroepitheliomas, rhabdomyocarcomas, and Ewing sarcomas [179], where in the latter plasma proCCK measurements have turned out to be a promising marker [180]. In the brain, acoustic neuromas [181], astrocytomas [182], and gliomas [183, 184] synthesize CCK peptide and express CCK receptors. So far, however, proper hyperCCKemia in humans has been recorded only in association with the mentioned metastatic pancreatic CCKoma [75, 176, 177].

It is always a question whether peptide hormones expressed in cell lines also mirror expression in the original tumors from which the cell lines are derived. Moreover, the amount and molecular form of the hormone as released in vivo from the original tumor also matters in terms of clinical phenotype. Nevertheless, beyond the transplantable pancreatic islet cell line [175], CCK gene expression at the peptide level has been encountered also in a human bronchial small‐cell carcinoma cell line [185] and a medullary thyroid carcinoma cell line [186].

Clinical perspectives: a summary

Considering the many different diseases in which CCK seems to be involved, much clinical research remains to be performed in order to establish the roles of CCK in pathophysiology, in diagnosis, and in therapy. The diseases mentioned in the present review are enumerated in Table 4. However, the research should always be carried out in awareness about the premise that the action of CCK on their target cells often will be a potentiating interaction with other messenger systems. Among such other systems, several will be other hormonal gastrointestinal peptides, other neuropeptide transmitters, or sometimes other cytokine peptides. That is complicated, but complexity is an inborn feature of the new biology.

Table 4.

Cholecystokinin in diseases.

Disorder	Role of CCK	Degree of evidence
Metabolic disorders
Obesity before bypass	None	+++
Obesity after bypass	↑ Incretin effect	++
Diabetes type 1	Exogenous CCK	–
Diabetes type 2	For incretin potentiation	++
Anorexia nervosa	None	+
Bulimia nervosa	None	+
Gastrointestinal disorders
Partial gastrectomy	↑ Secretion	+++
Sleeve gastrectomy	↑ Secretion	+++
Gastric bypass	↑ Secretion	+++
Colectomy	↑ Secretion	+++
Celiac disease	↓ Secretion	++
Pancreatic insufficiency	↓ Secretion	++
Fructose intolerance	↑ Secretion	+
Neuropsychiatric disorders
Depression	↓ CSF levels	+
Schizophrenia	↓ CSF levels	+
Parkinson's disease	↓ CSF levels	+
Multiple sclerosis	↑ CSF levels	+
Panic anxiety	Exogenous CCK‐4 provokes attacks	+++
Alzheimer's disease	Potential drug candidate
Cardiovascular disorders
Heart failure	ProCCK as prognostic marker	++
Neuroendocrine tumors
Pancreatic CCKoma	Diagnostic marker	++
Cushing tumors	Diagnostic marker	+
Nelson tumors	Diagnostic marker	+
Thyroid C‐cell cancer	Diagnostic marker	+
Pheochromocytomas	Diagnostic marker	+
Pediatric tumors
Askin tumors	Diagnostic marker	+
Neuroepitheliomas	Diagnostic marker	+
Rhabdomyosarcomas	Diagnostic marker	+
Ewing sarcomas	Diagnostic marker	++

With this background, some examples of obvious clinical applications can be proposed: In pharmacotherapy, new combination drugs should be developed. They should be based on a combination of CCK peptides (for instance, the CCK₂ receptor agonist nonsulfated CCK‐5) with other gut hormones such as GLP‐1 and perhaps GIP. The fraction of CCK‐5 should be sufficiently low to avoid damage in pancreatic exocrine cells, whereas CCK₂ receptor endocrine cells and neurons in the pancreas and the brain, respectively, could be activated and potentiate the already known effects of GLP‐1. Such drug(s) might improve therapy in both metabolic (obesity and type 2 diabetes mellitus [2]) and cerebral Alzheimer's and Parkinson's diseases.

In diagnostic oncology, it is described above that measurements of CCK or proCCK are useful markers for CCKomas and Ewing sarcomas. CCKoma is apparently a very rare neuroendocrine tumor. However, in order to ensure its prevalence and to learn more about the symptoms, larger series of plasma from neuroendocrine tumor cohorts in general and Zollinger–Ellison patients in particular [75] should be screened with reliable CCK assays as explained, discussed, and recommended in section V of this article (measurement of CCK). Furthermore, also plasma from patients with small‐cell lung carcinomas and thyroid cancer should be screened, because CCK peptides are expressed in cell lines from these tumors [185, 186].

In cardiology, there is still only little knowledge about the pathobiochemical and pathophysiological role of the proCCK fragments expressed in the cardial myocytes. The initial study of CCK in the heart [57] revealed that measurement of the cardiac proCCK fragment in plasma could be a prognostic risk factor for cardiac mortality. But there is a need for considerably more information, also about CCK receptor expression in the heart [187, 188].

In endocrinology, still nothing is known about plasma CCK levels in patients with pituitary corticotrophic tumors [178], thyroid C‐cell tumors [53], and pheochromocytomas [54]. Because these tumors express both bioactive CCK and CCK precursors, it is still an open question whether CCK peptides influence the pathophysiology and whether plasma CCK measurements might have diagnostic and/or prognostic significance.

Finally, in gastroenterology, the new recognition about CCK synthesis in the lower jejunum, the entire ileum, and also occasionally in the proximal colon [42] makes it worth examining the pre‐ and postprandial secretion of CCK in small intestinal diseases. For instance, the effects of fructose intolerance on CCK secretion might also be worth pursuing [189].

Author contributions

Jens F. Rehfeld: Writing—original draft; conceptualization; investigation; methodology; validation; visualization; writing—review and editing; software; formal analysis; project administration; data curation; supervision; resources.

Conflict of interest statement

The author declares no conflicts of interest.

Funding information

The writing of this article received no funding. Earlier cholecystokinin studies in the laboratory of the author, however, received support from the Danish Medical Research Council, the Danish Cancer Union, and the Novo Nordic Foundation.

Rehfeld JF. Cholecystokinin: Clinical aspects of the new biology. J Intern Med. 2025;298:251–267.

Data availability statement

Data sharing is not relevant for the present perspective article.