Ectodomain shedding of the hypoxia-induced carbonic anhydrase IX is a metalloprotease-dependent process regulated by TACE/ADAM17

Abstract

Carbonic anhydrase IX (CA IX) is a transmembrane protein whose expression is strongly induced by hypoxia in a broad spectrum of human tumours. It is a highly active enzyme functionally involved in both pH control and cell adhesion. Its presence in tumours usually indicates poor prognosis. Ectodomain of CA IX is detectable in the culture medium and body fluids of cancer patients, but the mechanism of its shedding has not been thoroughly investigated. Here, we analysed several cell lines with natural and ectopic expression of CA IX to show that its ectodomain release is sensitive to metalloprotease inhibitor batimastat (BB-94) and that hypoxia maintains the normal rate of basal shedding, thus leading to concomitant increase in cell-associated and extracellular CA IX levels. Using CHO-M2 cells defective in shedding, we demonstrated that the basal CA IX ectodomain release does not require a functional TNFalpha-converting enzyme (TACE/ADAM17), whereas the activation of CA IX shedding by both phorbol-12-myristate-13-acetate and pervanadate is TACE-dependent. Our results suggest that the cleavage of CA IX ectodomain is a regulated process that responds to physiological factors and signal transduction stimuli and may therefore contribute to adaptive changes in the protein composition of tumour cells and their microenvironment.

Figure 1

Basal carbonic anhydrase IX (CA IX) ectodomain shedding from CGL3 cells. (A) Schematic illustration of CA IX molecule with an extracellular portion composed of N-terminal proteoglycan (PG)-like region and central carbonic anhydrase (CA) domain linked via a single transmembrane segment to a short intracytoplasmic (IC) tail. Detection of the ectodomain was enabled by two noncompetitive CA IX-specific monoclonal antibodies: V/10 directed to CA and M75 directed to PG. The arrow indicates a putative cleavage site whose exact position is unknown. (B) CA IX ECD release was analysed by immunoprecipitation–immunoblotting analysis of the aliquots of nondiluted extracts (ex) and media (m) corresponding to the same number of cells obtained from the biotinylated CGL3 cultures that were chased for various time periods. CA IX was first precipitated with V/10 MAb, blotted and revealed with streptavidin–peroxidase. (C) The same blot was stripped and reprobed with the biotinylated M75 followed by streptavidin–peroxidase. (D) CA IX ECD release from biotinylated CGL3 cells was assessed also by sandwich ELISA using V/10 antibody as a capture and streptavidin–peroxidase as a detector. For this purpose, cell extracts were diluted 20 times, whereas media were diluted 1:2. OD₄₉₂ values represent average levels of CA IX measured in triplicates in at least three independent experiments, and standard deviations are indicated. (E) CGL3 cells grown on glass coverslips were allowed to detach in Ca²⁺-free PBS and then subjected to immunofluorescence labelling with M75 and FITC-conjugated anti-mouse antibodies. The cell nuclei were stained with propidium iodide. The arrow points to CA IX-positive ‘traces’ left after the detachment of CGL3 cell.

Figure 2

Basal rate of carbonic anhydrase IX (CA IX) ectodomain release from various cell lines under normoxia and hypoxia. (A) Confluent monolayers of CA IX-expressing cells were washed, fresh medium was added and the shedding was allowed to proceed over a 48-h period. The media were collected, the cells were extracted and the content of CA IX was determined by ELISA with V/10 as a capture and biotinylated M75 followed by streptavidin–peroxidase as a detector. (B) Serial dilutions of the extracts and media were examined by ELISA to relate the cellular and extracellular CA IX concentrations. (C) The parallel cultures were maintained for 48 h in normoxia and hypoxia (2% O₂). Harvested media (diluted 1:2) and cell extracts (diluted 1:20) were assayed in ELISA and the hypoxia-induced changes of CA IX levels in the cells and released to medium were illustrated as a ratio between the absorbance values measured in the hypoxic samples and the values of the corresponding normoxic samples. ^*P<0.01 and ^**P<0.001 by Student's t-test.

Figure 3

Inhibition of the basal carbonic anhydrase IX (CA IX) shedding with BB-94 metalloprotease inhibitor. (A) CGL3 cells were biotinylated and then allowed to shed CA IX ECD in the presence and absence of 1 μM BB-94. The media (m) were collected, the cells were extracted (ex) and the corresponding aliquots of the materials were analysed by immunoprecipitation–immunoblotting as described in Figure 1c. (B) ELISA was used to determine the inhibition of CA IX shedding with 1 μM BB-94. The cells were treated for 24 h throughout their incubation in normoxia or hypoxia. Extent of inhibition was illustrated as a ratio of the absorbance values of the materials from BB-94-treated cells vs their nontreated counterparts normalised according to total protein concentrations. (C) ELISA assessment of a BB-94 concentration-dependent inhibition of CA IX ECD release in MDCK+CA IX cells. The results are expressed as a ratio between the BB-94-inhibited and control samples. ^*P<0.01 and ^**P<0.001 by Student's t-test.

Figure 4

Activation of carbonic anhydrase IX (CA IX) ectodomain release by phorbol-12-myristate-13-acetate (PMA) and pervanadate. (A) The cells were first serum starved and then incubated with fresh medium in the presence or absence of 10% FCS, or alternatively in the presence or absence of epidermal growth factor (EGF), for 24 h. The extracts and media were analysed by ELISA and the results were illustrated as a ratio between the stimulated and nonstimulated cells. (B) The serum-starved cells were allowed to release CA IX ECD upon treatment with PMA for 24 h. The extracts and media were analysed by ELISA as above. (C) MDCK+CA IX cells were maintained overnight in serum-free medium and then treated for 2 h with PMA. Activation of CA IX shedding was revealed by immunoprecipitation–immunoblotting of the corresponding aliquots of the cell extracts and media similarly as in Figure 1. (D) MDCK+CA IX cells were serum-starved and then treated with PMA and/or pervanadate (V₂O₇) for 2 h. BB-94 was added 30 min before the induction. Collected media and cell extracts were assayed in ELISA and the results were illustrated as a ratio between the normalised absorbance data obtained from treated and control samples. (E) CGL3 cells grown on glass coverslips were serum-starved and then treated with PMA and pervanadate for 2 h. After allowing detachment of cells in Ca²⁺-free PBS, the coverslips were labelled with M75 followed by FITC-conjugated anti-mouse antibodies and the cell nuclei were stained with propidium iodide. The arrows indicate CA IX-positive ‘cell traces’. ^*P<0.01 and ^**P<0.001 by Student's t-test.

Figure 5

Expression of metalloproteases in selected human tumour cell lines and related collagenase activities. The cells were grown in serum-free media for 48 h. The media were collected for a collagen zymography and the cells were utilised as a source of RNA for RT–PCR. (A) Expression of metalloproteases was analysed by RT–PCR using the gene-specific primers listed in Table 1. β-Actin was used as a standard to monitor the amount of RNA templates. (B) Media were applied to a collagen-containing SDS–PAGE and collagenase activities of MMP1, -2 and -9 were evaluated according to an intensity of the nonstained zones representing a digested collagen. Medium from HT1080 fibrosarcoma cells was used as a positive control rich for MMP2 and -9.

Figure 6

Role of TACE/ADAM17 in the basal and regulated cleavage of carbonic anhydrase IX (CA IX) ECD. CHO-wild type (WT) cells, their mutated variant CHO-M2 defective in TACE/ADAM17 maturation, CHO-M2+TACE transfectants containing a fully functional TACE and CHO-Ka13 cells lacking HIF-1α were transiently transfected to express CA IX. (A) The transfected cells were allowed to shed CA IX into the medium over 24-h incubation period under normoxia and hypoxia (2% O₂). The same procedure was applied to mock-transfected cells. Media and extract collected from all dishes were assayed by ELISA. The results were expressed as average absorbance values with standard deviations. (B) The transfected cells were treated for 2 h either with PMA or with pervanadate in the absence and presence of BB-94 metalloprotease inhibitor added 30 min before the treatment. The results were illustrated as a ratio between the absorbance data obtained from the induced vs control samples. ^*P<0.01 and ^**P<0.001 by Student's t-test.