Calbindin28kDa is specifically associated with extranuclear constituents of the dense particulate fraction

Recent attempts to understand the function of calbindin28kDa, a widely expressed calcium-binding protein, are confounded by uncertainties over its subcellular location. Using immunoblot analysis of rat brain subregions, we found that the proportion of particulate calbindin28kDa (24-43% of total) was independent of expression level and location. The association of calbindin28kDa with particulate structures appeared to be specific, since it persisted when soluble calbindin28kDa was sequestered by antibodies added before tissue disruption. Moreover, when exogenous calbindin28kDa was added during homogenisation of brain from calbindin28kDa-nullmutant mice, only 10% partitioned to the particulate fraction compared with 33% of endogenous calbindin28kDa in wildtype controls. Confocal microscopy showed that calbindin28kDa was predominantly extranuclear in all tissues analysed (i.e. various brain regions, isolated neurons, and dental enamel epithelium). Dual-label microscopy of neural dense particulate fractions confirmed the extranuclear location of calbindin28kDa and also showed that it partly colocalised with synaptosome and microtubule markers. Using sucrose step gradients, calbindin28kDa was separated from nuclei in parallel with synaptosome and endoplasmic reticulum markers. However, no association with the marker proteins (synaptophysin, ERp29, alpha/beta-tubulin) was detected by calbindin28kDa-immunoprecipitation analysis. Together these findings provide the first consistent picture that calbindin28kDa is located predominantly outside of the nucleus, irrespective of tissue type (neuronal vs. non-neuronal) and experimental approach (biochemical vs morphological). The evidence of a substantial, strong and specific association with insoluble cellular structures challenges the widely held view of calbindin28kDa as a mobile calcium buffer, and supports the existence of important alternative roles that involve target proteins.     

Calbindin D-28k immunoreactive nerve fibers in the carotid body of normoxic and chronically hypoxic rats

The distribution and ultrastructural characteristics of calbindin D-28k immunoreactive nerve fibers were examined in the carotid body of the normoxic control rats by light and electron microscopy, and the abundance of calbindin D-28k fibers in the carotid body was compared in normoxic and chronically hypoxic rats (10% O2 and 3.0-4.0% CO2 for 3 months). Calbindin D-28k immunoreactivity was recognized in nerve fibers within the carotid body. Calbindin D-28k immunoreactive nerve fibers appeared as thin processes with many varicosities. They were distributed around clusters of glomus cells, and around blood vessels. Immunoelectron microscopy revealed that the calbindin D-28k immunoreactive nerve terminals are in close apposition with the glomus cells, and membrane specialization is visible in some terminals. Some dense-cored vesicles in the glomus cells were aggregated in this contact region. The chronically hypoxic carotid bodies were found to be enlarged several fold, and a relative abundance of calbindin D-28k fibers was lesser than in the normoxic carotid bodies. When expressed by the density of varicosities per unit area of the parenchyma, the density of calbindin D-28k fibers associated with the glomus cells in chronically hypoxic carotid bodies was decreased by 70%. These immunohistochemical findings indicate a morphological basis for involvement of calcium binding protein in the neural pathway that modulates carotid body chemoreception.     

Calcium Binding Protein in Squid Brain: Biochemical Similarity to the 28,000-Mr Vitamin D-Dependent Calcium Binding Protein (Calbindin-D28k)

A calcium binding protein that is biochemically similar to vertebrate 28,000-Mr vitamin D-dependent calcium binding protein (calbindin-D28k) has been purified from squid brain. Squid brain calbindin was found to have an isoelectric point of 5.0, was heat stable up to 60 degrees C, and showed increased electrophoretic mobility in the presence of chelator. Amino acid analysis revealed a high content of glutamic and aspartic acids and a low level of methionine, histidine, and tyrosine, a finding similar but not identical to the composition of vertebrate calbindin-D28k. The molecular weight of the squid protein, determined by Ferguson plot analysis of data obtained from sodium dodecyl sulfate-gel electrophoresis, was calculated to be 25,700, as compared with 27,800 for rat renal calbindin. Immunocytochemical analysis demonstrated immunoreactive protein in a selected population of neurons and fibers in several areas of the molluscan nervous system. This study represents the first purification from an invertebrate of a calcium binding protein that is biochemically similar to vitamin D-dependent calcium binding protein. These results demonstrate that calbindin, although not identical in vertebrates and cephalopods, may be phylogenetically conserved in structure. The restricted distribution of immunoreactive calbindin in both the cephalopod and mammalian brain suggests that the function of neuronal calbindin may also be conserved in evolution.     

Identification, purification, and partial characterization of a novel Mr 28,000 integral membrane protein from erythrocytes and renal tubules

A novel Mr 28,000 integral membrane protein ("28kDa") was identified in human erythrocytes and found entirely associated with the Triton X-100 insoluble membrane skeletons. Antibodies to 28kDa reacted strongly on immunoblots with 28kDa and a diffuse region of Mr 35,000-60,000 ("HMW-28kDa"). Selective proteolytic digestions of membranes demonstrated that HMW-28kDa has an extracellular domain, and both 28kDa and HMW-28kDa have intracellular domains. 28kDa and HMW-28kDa were purified to homogeneity. Quantitative immunoblots indicate that each erythrocyte contains 120,000-160,000 copies of 28kDa. Two-dimensional iodopeptide maps of 28kDa and HMW-28kDa were nearly identical; peptide-N-glycosidase digestion of purified HMW-28kDa demonstrated that it is the N-glycosylated form of 28kDa. When concentrated, 28kDa formed a series of larger oligomers which were stable in sodium dodecyl sulfate. Of several nonerythroid tissues studied with anti-28kDa immunoblots, only kidney displayed immunoreactive 28kDa. Purified rat kidney 28kDa was nearly identical to rat erythrocyte 28kDa when compared by two-dimensional iodopeptide mapping. Immunohistochemical staining of human kidney with anti-28kDa demonstrated prominent staining over the apical brush borders of proximal convoluted tubules. A novel integral membrane protein has been purified from erythrocyte and kidney membranes. This new protein may play a role in linkage of the membrane skeleton to the lipid bilayer.     

The human genes for calbindin 27 and 29 kDa proteins are located on chromosomes 8 and 16, respectively

Genomic clones coding for the brain calcium-binding protein, calbindin 29 kDa, were isolated from a human library. A fragment containing exon 2 was used as a probe to investigate the presence of the gene in human x rodent somatic cell hybrids. The gene was unambiguously assigned to chromosome 16. The closely-related calbindin 27 kDa gene was previously assigned to chromosome 8. These two genes, deriving from a common ancestor, thus appear to have been separated during vertebrate evolution.     

Distribution and co-localization of calbindin D28k with VIP and neuropeptide Y but not somatostatin, galanin and substance P in the enteric nervous system of the rat

Calbindin D28k, previously demonstrated in the mammalian central nervous system, has been localized to discrete neurons in the enteric nervous system of the rat. Calbindin D28k is present in cell bodies in both the myenteric and submucous plexi and in interganglionic nerve fibers in all regions of the gastrointestinal tract. Immunoreactive nerve fibers were also detected in the mucosal region, although none were observed in the pyloric sphincter, circular or longitudinal muscle layers. The highest concentration of immunoreactivity was present in the submucosal plexus and mucosa of the colon. Western blot analysis of the protein detected by the antiserum confirmed that it comigrated with purified calbindin D28k and the single immunoreactive band seen in extracts from rat brain. The colocalization of calbindin D28k with components of the peptidergic innervation was also investigated. Of the peptides studied the neurons containing both vasoactive intestinal polypeptide and neuropeptide Y in the submucous plexus were seen to exhibit calbindin D28k immunoreactivity. The neurons containing somatostatin, galanin and substance P did not demonstrate co-localization. In the stomach, calbindin D28k was detected within a small number of epithelial cells which were found to correspond to a sub-population of the somatostatin-immunoreactive endocrine cells.     

Expression in Escherichia coli of full-length and mutant rat brain calbindin D28. Comparison with the purified native protein

Studies of vitamin D-dependent 28-kilodalton calcium binding protein (calbindin D28) have been hindered by difficulties in purifying large amounts of the protein. In order to overcome this problem, we cloned and expressed a full-length rat brain calbindin D28 cDNA. In addition, we isolated and purified to homogeneity, native rat brain calbindin D28. The isolated native protein has an apparent molecular mass of 27 kDa and properties similar to those of the well-characterized chicken calbindin D28. It has an acidic isoelectric point (approximately 4.5), a high affinity for calcium, and an amino terminus blocked to Edman degradation. The properties of the native and the recombinant proteins were examined by gel electrophoresis, isoelectric focusing, protein sequencing, amino acid composition analysis, and calcium binding assays. We demonstrated that: (i) the authentic and the full-length recombinant proteins have similar molecular weights and isoelectric points; (ii) the proteins have the same amino acid composition; (iii) the proteins bind calcium in a similar manner; (iv) the absence of a blocking NH2-terminal group in the recombinant protein does not appreciably influence the binding of calcium. To further examine the calcium binding properties of this protein, we constructed deletion mutants lacking one or both of the two putative degenerated calcium binding sites (EF hand regions). These deletions resulted in smaller proteins that still bound calcium. The ability to express and purify calbindin D28 and mutants thereof should allow the systematic elucidation of structure-function relationships in this class of calcium binding proteins.     

Calretinin and calbindin in the retina of the developing chick

Summary Calretinin and calbindin-D28k are two calcium-binding proteins that are present in largely different sets of nerve cells in the central nervous system. Their appearance during development of the chick retina was studied by immunohistochemistry and Western blots. The patterns are mature one day before hatching. Each cell type acquires its characteristic calcium-binding protein several days after its differentiation has started, but in most cases before morphological maturation is complete. There is also an early phase of calbindin immunoreactivity in many immature amacrine cells, and of calretinin immunoreactivity in the presumptive photoreceptor layer, suggesting that these proteins may have distinct functions in differentiating cells.Abbreviations CR+ Immunoreactive for calretinin only - CB+ immunoreactive for calbindin only - CR+CB+ immunoreactive for both antisera - IPL inner plexiform layer - OPL outer plexiform layer     

Localization of calbindin D28K-like immunoreactivity in fish gill: a light microscopic and immunoelectron histochemical study.

The presence of calbindin D28K in fish (Heteropneustes fossilis) gill was studied by use of specific antibodies raised against chick duodenal 28 kDa calbindin in immunoperoxidase and electron-microscopic labelling experiments. Immunoreactivity for calbindin D28K, which has been observed in the intestine of a number of avian and mammalian species, is reported for the first time in the gill. It was primarily located in neuroendocrine (NE) cells. Some immunoreactivity was also located in the glycocalyx of the non-endocrine cells, i.e., the pavement cells, which have ultrastructural characteristics quite different from those of endocrine cells. The calbindin-immunopositive NE cells were ascertained in both gill filamental and lamellar epithelium. All the NE cells contained secretory granules as the most distinctive feature of these cells. Ultrastructurally, two types of NE cells were distinguished according to the morphology of their secretory granules. The calbindin immunoreactivity in the NE cells was stimulated when the calcium concentration of the ambient water was reduced. The present findings suggest that NE cells exert some as yet unidentified function related to calcium-mediated processes involving the expression of calbindin.     

Separate neurochemical classes of sympathetic postganglionic neurons project to the left ventricle of the rat heart

The sympathetic innervation of the rat heart was investigated by retrograde neuronal tracing and multiple label immunohistochemistry. Injections of Fast Blue made into the left ventricular wall labelled sympathetic neurons that were located along the medial border of both the left and right stellate ganglia. Cardiac projecting sympathetic postganglionic neurons could be grouped into one of four neurochemical populations, characterised by their content of calbindin and/or neuropeptide Y (NPY). The subpopulations of neurons contained immunoreactivity to both calbindin and NPY, immunoreactivity to calbindin only, immunoreactivity to NPY only and no immunoreactivity to calbindin or NPY. Sympathetic postganglionic neurons were also labelled in vitro with rhodamine dextran applied to the cut end of a cardiac nerve. The same neurochemical subpopulations of sympathetic neurons were identified by using this technique but in different proportions to those labelled from the left ventricle. Preganglionic terminals that were immunoreactive for another calcium-binding protein, calretinin, preferentially surrounded retrogradely labelled neurons that were immunoreactive for both calbindin and NPY. The separate sympathetic pathways projecting to the rat heart may control different cardiac functions.     

Identification of N-acetylgalactosamine-containing glycoproteins PEB3 and CgpA in Campylobacter jejuni

It was demonstrated recently that there is a system of general protein glycosylation in the human enteropathogen Campylobacter jejuni. To characterize such glycoproteins, we identified a lectin, Soybean agglutinin (SBA), which binds to multiple C. jejuni proteins on Western blots. Binding of lectin SBA was disrupted by mutagenesis of genes within the previously identified protein glycosylation locus. This lectin was used to purify putative glycoproteins selectively and, after sodium dodecyl sulphatepolyacrylamide gel electrophoresis (SDS-PAGE), Coomassie-stained bands were cut from the gels. The bands were digested with trypsin, and peptides were identified by mass spectrometry and database searching. A 28kDa band was identified as PEB3, a previously characterized immunogenic cell surface protein. Bands of 32 and 34kDa were both identified as a putative periplasmic protein encoded by the C. jejuni NCTC 11168 coding sequence Cj1670c. We have named this putative glycoprotein CgpA. We constructed insertional knockout mutants of both the peb3 and cgpA genes, and surface protein extracts from mutant and wild-type strains were analysed by one- and two-dimensional polyacrylamide gel electrophoresis (PAGE). In this way, we were able to identify the PEB3 protein as a 28 kDa SBA-reactive and immunoreactive glycoprotein. The cgpA gene encoded SBA-reactive and immunoreactive proteins of 32 and 34 kDa. By using specific exoglycosidases, we demonstrated that the SBA binding property of acid-glycine extractable C. jejuni glycoproteins, including PEB3 and CgpA, is a result of the presence of alpha-linked N-acetylgalactosamine residues. These data confirm the existence, and extend the boundaries, of the previously identified protein glycosylation locus of C. jejuni. Furthermore, we have identified two such glycoproteins, the first non-flagellin campylobacter glycoproteins to be identified, and demonstrated that their glycan components contain alpha-linked N-acetylgalactosamine residues.     

Alterations in the localization of calbindin D28K-, calretinin-, and parvalbumin-immunoreactive neurons of rabbit retinal ganglion cell layer from ischemia and reperfusion

Calcium-binding proteins are thought to play important roles in calcium buffering. The present study investigated the effects of ischemia and reperfusion on calbindin D28K, calretinin, and parvalbumin immunoreactivity in the ganglion cell layer of the rabbit. Rabbits were administered ischemic damage by increasing the intraocular pressure. After 60 and 90 min of ischemia, reperfusion (7 d) was allowed to occur. The b-wave of the electroretinogram (ERG) was reduced by more than 50% and almost 80% in retina given ischemia for 60 and 90 min, respectively. The oscillatory potential (OPs) wave was reduced approximately 50% at 60 min ischemia and 70% at 90 min ischemia. In both normal and ischemic-treated retina, calcium-binding protein immunoreactivity was seen in many cells in the ganglion cell layer. In eyes subjected to 60 min ischemia, there was a decrease of the density of calbindin D28K- (8.29%), calretinin- (14.44%), and parvalbumin- (26.83%) immunoreactive (IR) cells compared to the control retina. In eyes subjected to 90 min ischemia, there was a higher decrease of the density of calbindin D28K- (18.48%), calretinin- (33.59%), and parvalbumin- (54.26%) IR cells than at 60 min. Some calcium-binding protein-IR neurons, especially calretinin-IR neurons, showed aggregations that were abnormally packed together in retina subjected to ischemia for 90 min. The results show that calbindin D28K-, calretinin-, and parvalbumin-IR cells in the ganglion cell layer are susceptible to ischemic damage and reperfusion. The degree of reduction varied among different calcium-binding proteins and ischemic damage times. These results suggest that calbindin D28K-containing neurons are less susceptible to ischemic damage than calretinin- and parvalbumin-containing neurons in the ganglion cell layer of rabbit retina.     

An immunoreactive form of erythrocyte protein 4.9 is present in non-erythroid cells

Using immunoblots and an affinity-purified antibody prepared against human erythrocyte protein 4.9, we have demonstrated and quantified the presence of an immunoreactive form of this protein in avian and bovine brain and lens tissues, avian heart, as well as in human platelets and mammalian, avian, piscine, and amphibian erythrocytes. Both the 48 kDa and the 52 kDa variants were observed in human erythrocytes, whereas 50 kDa and 54 kDa immunoreactive forms were observed in human platelets. As reported for erythroid protein 4.9, platelet protein 4.9 was phosphorylated in response to treatment with phorbol ester. Bovine brain showed five cross-reactive polypeptides in the 47 to 52 kDa range while avian brain and avian and bovine lens exhibited predominantly a 49-kDa band. Cross-reactivity was not observed in a number of cell lines and tissues including leukocytes, liver, kidney, pancreas, and skeletal muscle. Immunofluorescence indicated that protein 4.9 was present in cortical fiber cells of avian lens and in neurons of avian cerebrum.     

Presence of oxidized protein hydrolase in human cell lines, rat tissues, and human/rat plasma

Oxidized protein hydrolase (OPH), an 80 kDa serine protease whose activity is inhibited by diisopropyl fluorophosphate (DFP), has been isolated from human erythrocytes [Fujino, T. et al. (1998) J. Biochem. 124, 1077-1085]. The presence of OPH in various biological samples was examined by enzyme-linked immunosorbent assay (ELISA) and immunoblotting using an anti-OPH antibody raised against OPH purified from human erythrocytes, and by [(3)H]DFP-labeling and successive SDS-PAGE/fluorography. Solubilized samples of human cell lines including K-562 cells, THP-1 cells and Jurkat cells, and rat tissues including brain, heart, liver, kidney, and testis, inhibited the anti-OPH antibody binding to OPH in ELISA. Immunoblotting of lysates of K-562 cells, THP-1 cells and Jurkat cells showed four immunoreactive protein bands including an 80 kDa protein. Immunoprecipitation of the [(3)H]DFP-labeled K-562 cell lysate and successive SDS-PAGE/fluorography showed the presence of only the 80 kDa DFP-reactive protein with OPH antigenic activity. The level of the 80 kDa immunoreactive protein in K-562 cells rose as the cells differentiated toward erythrocytes. Immunoblotting of human and rat plasma showed two immunoreactive protein bands, including the 80 kDa protein, and SDS-PAGE/fluorography of [(3)H]DFP-labeled rat and human plasma showed the presence of only the 80 kDa DFP-reactive protein. The results indicate that OPH is present in a wide variety of biological samples.     

Calbindin D-28K in the nucleus pretectalis superficialis parvicellularis of Salmo gairdneri: an immunocytochemical study.

The distribution of calbindin D-28K (CaBP28K) cell bodies and fibers in the nucleus pretectalis superficialis parvicellularis of the rainbow trout was studied using a monoclonal antibody and the avidin-biotin-peroxidase method. In this diencephalic nucleus a very high density of CaBP28K immunoreactive fibers was found. In addition, a high density of CaBP28K positive neurons was also observed. These neurons were small, showing one, two or three short and non-branching dendritic trunks. The distribution and orientation of the immunoreactive cell bodies in the nucleus pretectalis superficialis parvicellularis suggests that the neurons might be interneurons and/or projecting neurons.     

A Folate Binding Protein in Ascitic Fluid, Serum and Ovarian Tissue of Patients with Ovarian Adenocarcinoma Immunoreacts with Antibodies Against Human Milk Folate Binding Protein

The presence of a folate binding protein which immunoreacts with antibodies against human milk folate binding protein was demonstrated in ascitic fluids from seven patients with ovarian adenocarcinoma. Ascitic fluids collected from two patients with other malignancies contained non-immunoreactive FBP. Tumor tissue specimens from five patients with ovarian carcinoma contained immunoreactive FBP. By contrast to normal ovaries ovarian carcinoma tissue showed positive immunostaining on immunohistochemistry. Ascitic fluids from two patients with ovarian carcinoma exhibited single distinct bands on SDS-PAGE immunoblotting. The gel filtration profile of ovarian carcinoma tissue homogenate from two patients contained 25 and 100 kDa peaks of radioligand-bound and immunoreactive folate binding protein, while ascitic fluid from one of the patients exhibited a large 100 kDa immunoreactive peak with no radioligand binding activity. The immunoreactive non-functional 100 kDa FBP could represent unprocessed precursor FBP. Future studies are necessary to evaluate whether determination of immunoreactive FBP in ovarian adenocarcinomatosis is of any diagnostic value.     

A high-affinity folate binding protein in human semen

The presence of a folate binding protein of high-affinity type (affinity constant 3.10¹⁰M^–1, maximum folate binding 1.4 nM) in human semen was demonstrated in equilibrium dialysis experiments (37°C, pH 7.4) with the radioligand³H-folate. Radioligand dissociation from the binding protein was slow at pH 7.4, but rapid at pH 3.5. By use of rabbit antibodies against 25 kDa human milk folate binding protein we determined the concentration of folate binding protein in 16 speciments of human semen in an enzyme-linked immunosorbent assay. The concentration of immunoreactive folate binding protein was independent of the number of spermatozoa in individual specimens. Gel filtration showed that immunoreactive and radioligand bound folate binding protein coeluted in two peaks: a major one of 100 kDa and a minor one of 25 kDa.     

The carboxy-terminal extension of the collagen binding domain of fibronectin mediates interaction with a 165 kDa membrane protein involved in odontoblast differentiation

Terminal differentiation of the odontoblast is characterized by an elongation and a polarization of the cell. The change in the cell shape and the reorganization of the cytoplasm involve the microfilament system. An immunological approach has previously implicated a transmembrane interaction between fibronectin and vinculin in the control of odontoblast differentiation. A 165 kDa protein localized on the cell-surface of odontoblasts mediated this interaction. In order to define the nature of the interaction of the 165 kDa protein with fibronectin, peptides were prepared by proteolytic cleavage of fibronectin with alpha-chymotrypsin. The results indicate that the 165 kDa protein interacted with a 62 kDa peptide located towards the amino-terminal extremity of fibronectin, but not with a 47 kDa related fragment. Both these 62 kDa and 47 kDa peptides included the collagen-binding domain and were retarded on a heparin-Ultrogel column. Microsequences demonstrated that the 62 kDa and 47 kDa fragments had the same amino-terminal extremity and that the larger fragment was extended in the carboxy-terminal direction. This carboxy-terminal extension of the collagen binding domain of fibronectin is implicated in the interaction of this molecule with the 165 kDa protein. On the other hand, odontoblasts differentiated normally when tooth germs were cultured in the presence of GRGDS synthetic peptide, suggesting that RGD-dependent integrins were not involved in odontoblast differentiation. Staining of dental mesenchymal cells in primary culture and of differentiated odontoblasts in situ with antibodies directed against the beta 1-subunit of integrins confirmed previous observations and showed that although beta 1 integrins are involved in the attachment of cultured dental cells, they are not implicated in the process of odontoblast differentiation.     

Expression of brush border calmodulin-binding proteins during human small and large bowel differentiation

The expression and immunocytochemical localization of three brush border cytoskeletal calmodulin-binding proteins, caldesmon, fodrin, and the 110 kDa subunit of the 110 kDa calmodulin complex, have been studied in human intestinal epithelial cells as a function of their ontogenic differentiation. At immature stages (fetal week 8), caldesmon and fodrin were present in undifferentiated intestinal epithelial cells. However, no 110 kDa protein was detectable except a 135 kDa immunoreactive species. The 110 kDa form appeared at week 12, when microvilli differentiate, and became prominent at week 14 simultaneously with the disappearance of the 135 kDa species. Finally at week 14, the calmodulin-binding protein pattern was identical to that found in adults. Immunocytochemical experiments revealed that at week 8, antibodies to caldesmon and fodrin gave a fluorescence lining at the periphery of the cells, whereas the 110 kDa immunoreactive species was hardly detectable. Then, as early as week 12 of gestation, with the three antisera, a bright fluorescence lined the apex of the cells, as in adults. In the colon, the events were delayed. This study demonstrates that the developmental pattern of the three calmodulin-binding proteins investigated, caldesmon, fodrin and the 110 kDa subunit, parallels the temporal differentiation of human intestinal brush borders and the proximo-distal morphological intestinal maturation.     

Calbindin independence of calcium transport in developing teeth contradicts the calcium ferry dogma

Cytosolic calcium-binding proteins termed calbindins are widely regarded as a key component of the machinery used to transport calcium safely across cells. Acting as mobile buffers, calbindins are thought to ferry calcium in bulk and simultaneously protect against its potentially cytotoxic effects. Here, we contradict this dogma by showing that teeth and bones were produced normally in null mutant mice lacking calbindin(28kDa). Structural analysis of dental enamel, the development of which depends critically on active calcium transport, showed that mineralization was unaffected in calbindin(28kDa)-null mutants. An unchanged rate of calcium transport was verified by measurements of (45)Ca incorporation into developing teeth in vivo. In enamel-forming cells, the absence of calbindin(28kDa) was not compensated by other cytosolic calcium-binding proteins as detectable by (45)Ca overlay, two-dimensional gel, and equilibrium binding analyses. Despite a 33% decrease in cytosolic buffer capacity, cytotoxicity was not evident in either the null mutant enamel or its formative cells. This is the first definitive evidence that calbindins are not required for active calcium transport, either as ferries or as facilitative buffers. Moreover, in challenging the broader notion of a cytosolic route for calcium, the findings support an alternative paradigm involving passage via calcium-tolerant organelles.