Oncogene (2010) 29, 6509–6521 & 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 www.nature.com/onc REVIEW New insights into the physiological role of IX in tumour pH regulation

P Swietach1, A Hulikova1, RD Vaughan-Jones1,3 and AL Harris2,3

1Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, UK and 2Weatherall Institute of Molecular Medicine, John Radcliffe Hospital, University of Oxford, Oxford, UK

In this review, we discuss the role of the tumour-associated acidity, commonly expressed using the pH scale. A carbonic anhydrase isoform IX (CAIX) in the context of particularly important class of pH-sensitive molecules is pH regulation. We summarise recent experimental find- protein because of the strong link between protonation ings on the effect of CAIX on cell growth and survival, and state, tertiary structure and functional output. The þ present a diffusion-reaction model to help in the assess- responsiveness to pH depends on H affinity (KH), ment of CAIX function under physiological conditions. bestowed on the protein by the chemistry of its amino CAIX emerges as an important facilitator of acid residues. For many proteins, KH is numerically diffusion and acid transport, helping to overcome large close to physiological [H þ ], thus even small displace- cell-to-capillary distances that are characteristic of solid ments from ‘normal’ intracellular pH (pHi) can sig- tumours. The source of substrate for CAIX catalysis is nificantly alter cellular biochemistry. H þ ions serve as likely to be CO2, generated by adequately oxygenated universal and potent modulators of virtually all aspects mitochondria or from the titration of metabolic with of life. Through the breadth of their effects, H þ ions are À HCO3 taken up from the extracellular milieu. The appropriate signalling molecules for coordinating com- relative importance of these pathways will depend on plex cell programmes. For instance, moderately alkaline oxygen and metabolite availability, the spatiotemporal pHi is permissive for cell proliferation (Pouyssegur et al., patterns of the cell’s exposure to hypoxia and on the 1984; Chambard and Pouyssegur, 1986; Gillies et al., regulation of metabolism by . This is now an 1990; Gillies et al., 1992). It is not surprising that proper important avenue for further investigation. The impor- pHi regulation is of fundamental importance to all cells. tance of CAIX in regulating tumour pH highlights the Specialised mechanisms for pHi regulation have protein as a potential target for cancer therapy. evolved to fine-tune acid/base balance appropriately Oncogene (2010) 29, 6509–6521; doi:10.1038/onc.2010.455; for on-going cellular physiology. A major challenge to published online 4 October 2010 pHi house keeping is the large disparity between the target H þ ion concentration, typically between 50 and Keywords: cancer; pH; carbonic anhydrase IX; acid/ 100 nM, and the acid/base challenges, as high as base transport; buffers millimolar per minute, that disturb pHi. Among these acid/base loads is cellular respiration, which produces þ H -yielding CO2 or lactic acid. The most metabolically active cells are expected to express the most powerful The fundamental importance of acid/base balance pHi regulating apparatus. Cancer cells, for instance, are characterised by a very high metabolic rate (Gatenby Cellular function, growth and development arise from a and Gillies, 2004). The alkaline pHi measured in diversity of interlinked biochemical processes. The tumours with nuclear magnetic resonance (Griffiths proper orchestration of these unitary events is per- et al., 1981; Gillies et al., 2004) suggests that these cells formed by a hierarchy of signalling cascades, involving are well equipped to deal with excess H þ ions. both small ions (such as calcium) as well as more complex molecules (such as G-proteins). The potency and scope of these signals depends on the responsiveness of target molecules. Many biological molecules are weak Buffering reactions and membrane transport regulate acids or weak bases that can release or bind H þ ions intracellular pH (protons), respectively. The protonation state (and hence charge) of these molecules will depend on ambient Cells contain a high concentration of weak acids and weak bases—such as proteins, amino acids and phos- Correspondence: Dr P Swietach, Department of Physiology, Anatomy phates—which results in the intracellular milieu having and Genetics, University of Oxford, Sherrington Building, South Parks a high pH buffering power (Boron, 2004). Buffers are Road, Oxford OX1 3PT, UK. necessary to ‘dampen’ pHi displacements in response to E-mail: [email protected] þ 3Joint senior authors. H production (or consumption) that takes place at Received 16 July 2010; revised 30 August 2010; accepted 31 August 2010; distinct loci within cells, such as mitochondria or the published online 4 October 2010 surface membrane. In the presence of buffers, even Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6510 millimolar concentrations of acid (or base) inflict merely Therefore, pHi sensitivity must be built into the nanomolar changes in free H þ ion concentration apparatus for regulating acid/base balance. The fidelity (equivalent to changing pHi by a fraction of a unit). of this pHi-sensing mechanism relies on good diffusive Most cells usually undergo periods of sustained acid coupling across the cell, particularly between the bulk loading. Buffering does not eliminate the acid/base cytoplasm and the . The diffusive H þ flux problem, but merely acts as a temporary repository for must match H þ production with H þ extrusion (and H þ ions. Chronic acid production (for example, by similarly, H þ depletion with H þ loading). However, respiration) can deplete protonatable buffer sites. To because of the high intracellular buffering capacity, avoid this shortfall in buffering power, excess H þ ions, most H þ ions are bound to buffer molecules and can above a certain desired [H þ ] level, must be extruded diffuse only as fast as the H þ buffer complex (Irving from cells. In its simplest case, acid extrusion should et al., 1990; Vaughan-Jones et al., 2002; Swietach et al., match metabolic acid production. High-charge density 2003). An indispensible role for smaller buffer mole- precludes H þ ions themselves from crossing the lipid cules, such as amino acids, dipeptides, phosphates and À bilayer of membranes. To overcome this permeability /bicarbonate (CO2/HCO3), is to shuttle obstacle, H þ ions must be shuttled across the membrane H þ ions within cells (Figure 1a). These are needed to in a permeant form by transport proteins. As the offset the diffusive restrictions imposed by heavier H þ transport of H þ ions across the membrane may require buffers, such as proteins. an input of energy, pHi regulation usually involves An alternative buffer/transporter arrangement for þ À primary (for example, H ATPase) or secondary (for pHi regulation relies uniquely on the CO2/HCO3 buffer example, Na þ /H þ exchanger (NHE)) active transpor- system (Figure 1b). This buffer, often labelled as ters (Figure 1a; Martinez-Zaguilan et al., 1993; Lee and physiological, is a major component of buffering Tannock, 1998; McLean et al., 2000; Boron, 2004; capacity (Leem et al., 1999; Boron, 2004; Alper, 2006). Chiche et al., 2010). Unlike many other buffer systems, its protonated form Acid extrusion must remain responsive to even the (CO2) is usually freely permeant across the membrane smallest change in metabolic rate. Moreover, under lipid bilayer. Transmembrane flux of CO2 is therefore a certain circumstances, cells may incur a deficiency of H þ disguised form of H þ flux that does not require a ions and require acid uptake (for example, by ClÀ/H þ - protein transporter, although it can be enhanced by equivalent transport; Vasseur et al., 1989; Sun et al., aquaporins (Nakhoul et al., 1998). To enable this system 1996; Niederer et al., 2008), rather than acid extrusion. to regulate pHi effectively, flux of CO2 must be matched À by counter flux of HCO3. Bicarbonate ions are À membrane impermeant and therefore active HCO3 transport at the cell membrane is required to maintain À intracellular CO2/HCO3 composition (Figure 1b). For the purpose of this review, the two schemes of À pHi regulation are referred to as HCO3 dependent and À HCO3 independent with reference to the species of ion that is transported across the cell membrane. It is À noteworthy that, even in the absence of HCO3 À transporters, CO2/HCO3 buffer can participate in pH regulation by spatially shuttling H þ ions to and from H þ transporters (Spitzer et al., 2002; Swietach et al., 2007). À The versatility of the CO2/HCO3 buffer system can be limited by the slow kinetics of CO2 hydration and the reverse dehydration reaction. Without catalysis, the À slow interchange between CO2 and HCO3 could impede H þ shuttling within the cell and across the membrane. For this reason, most cells express carbonic anhydrase (CA) , which catalyse the reversible reaction, À þ CO2 þ H2O2HCO3 þ H . Some CA isoforms have an Figure 1 Cellular mechanisms for removing intracellular acid intracellular catalytic site (for example, CAI, CAII, þ À involve pH buffers and membrane H or HCO3 transporters. (a) CAIII and so on), whereas others (for example, CAIV, À þ HCO3-independent mechanism: Intracellular H ions are extruded CAIX, CAXII, CAXIV) catalyse the reaction at the across the cell membrane by means of an H þ transporter (hexagon), such as Na þ /H þ exchange or H þ lactate cotransport. extracellular surface of the cells (Chegwidden et al., The transporter is supplied with H þ ions by protonated buffers 2000; Supuran, 2008). þ À (HBuf) that facilitate intracellular H diffusion. (b) HCO3- Regulation of pHi goes beyond the cell. At tissue dependent mechanism: intracellular H þ ions are buffered by level, acid/base transport at the cell membrane must be À À HCO3 ions that are imported into cells by means of a HCO3 coupled diffusively with the blood inside capillaries. In transporter (pentagon), such as Na þ /HCOÀ co-transport. The 3 health, capillary blood plasma is well buffered and efflux of the membrane-permeant acid product, CO2, represents acid removal from the cell. Similar fluxes working in reverse can normally held close to pH 7.4. Continuous blood produce acid loading to correct high intracellular pH. perfusion tends to clamp the pH of the extracellular

Oncogene Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6511 fluid surrounding cells to 7.4. However, within inade- quately perfused tissues, as happens in many tumours (Kallinowski et al., 1989; Vaupel et al., 1989), extra- cellular pH (pHe) gradients develop as a result of long diffusion distances. Consequently, the acid-releasing cells that are more distant from capillaries will be surrounded by a milieu of lower pHe. As many membrane transporters are sensitive to pHe, this may feed back on acid/base traffic carried by such transpor- ters (Vaughan-Jones and Wu, 1990; Sun et al., 1996; Stewart et al., 2009). Homeostasis of pH in tissue must therefore involve processes that regulate pHi and pHe.

CAIX is an hypoxia-inducible, extracellular facing CA isozyme found in many tumours

The microenvironment of solid tumours is shaped by cancer biochemistry and physiology. As a prerequisite for colonising body resources, tumours grow at a rate that often exceeds the capacity of the host vasculature. Inadequate blood supply would normally impede further development of most tissues. Tumours, however, escape this negative feedback by maintaining high respiratory rates and surviving, even when oxygen, a principal blood-borne substrate, is scarce. Indeed, low oxygen tension (hypoxia) is a cardinal feature of the tumour milieu (Gatenby and Gillies, 2004). Hypoxia is not only a consequence of tumour respiration, but also a trigger for an altered programme of expression, featuring hypoxia-inducible genes (Gleadle and Rat- cliffe, 1998; Harris, 2002; Pouyssegur et al., 2006) that progress cancer towards a more aggressive disease Figure 2 CAIX is a dimeric membrane-tethered with extracellular carbonic anhydrase activity. (a) Fixed monolayer of phenotype (Fang et al., 2008). Many of the hypoxia- HCT116 cells transfected with the ca9 gene show CAIX protein regulated genes are controlled by hypoxia-inducible staining at the cell membrane (antibody raised against human factor, a transcription factor that is otherwise inacti- CAIX). (b) CAIX structure based on the X-ray crystal structure of vated in the presence of oxygen. The targets of hypoxia- the CAIX extracellular domain from the RCSB Protein Data Bank inducible factor include genes that encode for proteins (data from Alterio et al. (2009), rendered with Jmol viewer), showing the dimeric structure of the protein and the four distinct involved in glucose metabolism, blood vessel growth, pairs of domains: the proteoglycan domain (PG), catalytic domain oxygen carriage, iron metabolism and numerous other (CD), transmembrane segment (TM) across the membrane (M) and processes. the intracellular tail (IC). (c) The catalytic activity of CAIX was In the context of pH regulation, it has emerged that tested in CAIX over-expressing HCT116 cells, labelled with a pH- the enzyme CAIX, cloned originally in the mid 1990s sensitive dye at the extracellular surface of the membrane. Surface pHe (average of six cells) was measured during exposure to NH4Cl- À (Pastorek et al., 1994), is inducible by hypoxia (Wykoff containing solution, under superfusion with CO2/HCO3 buffer. et al., 2000). CAIX expression, unlike most other CA Transient surface pHe displacements from bulk solution pH (of 7.2) isoforms, is associated with many tumours (Pastorek were produced in response to NH3 influx and efflux, as explained in the cartoons. With CAIX activity, surface pHe transients were et al., 1994; Pastorekova and Pastorek, 2004; De Simone À small, because of the efficient buffering by CO2/HCO3. In the and Supuran, 2010). Indeed, very few normal tissues presence of 100 mM acetazolamide (ATZ), a broad spectrum CA (with the notable exception of stomach; Pastorekova inhibitor, surface pHe were considerably larger because of slower À et al., 1997) express significant levels of CAIX, so that CO2/HCO3 buffering kinetics in the absence of CA activity. positive staining for CAIX is now an established marker of tumour hypoxia and a clinical indicator of aggressive cancers (for example, breast and bone) with poor prognosis (Chia et al., 2001; Giatromanolaki et al., lar carboxy terminal tail that may be involved in cell–cell 2001; Generali et al., 2006). CAIX is membrane tethered adhesion and in regulating the catalytic process (Zavada (Figure 2a) and its catalytic domain is extracellular et al., 2000; Svastova et al., 2003; Hulikova et al., 2009). facing. According to a recent crystal structure Recent work (Innocenti et al., 2009) has proposed that (Figure 2b), CAIX exists as a dimer (Alterio et al., the presence of the proteoglycan domain, rich in acidic 2009; De Simone and Supuran, 2010). The protein amino acid residues, reduces the inhibitory effects of H þ contains a proteoglycan-like domain and an intracellu- ions on CAIX activity. This is observed as a shift in the

Oncogene Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6512 pH sensitivity of CAIX activity by half a pH unit metabolon (Sterling et al., 2001; Morgan et al., 2007). towards more acidic values, enabling CAIX to remain Such binding has been questioned (Boron, 2010), but catalytically active in the acidic extracellular milieu that even in the absence of direct binding, functional is typical of solid tumours. coupling may still occur. The focus of this review is on the physiology of the catalytic domain of the enzyme.

CA can facilitate acid removal in diffusionally restricted spaces À CAs reduce out-of-equilibrium episodes of CO2/HCO3 buffer Experimental solutions and blood plasma are large À reservoirs of equilibrated CO2/HCO3 buffer at stable CA activity accelerates the attainment of equilibrium pH. If diffusive coupling between cells and these þ À À between H ,CO2 and HCO3. Expressed mathemati- reservoirs is good, then CO2/HCO3 buffer is likely to þ À þ À cally, this tends to bring [H ] Â [HCO3]/[CO2] towards be at equilibrium. Diffusion of H ,CO2 or HCO3 to or the equilibrium constant of the buffer (K ¼ 10À6.2 Mat from the cell surface would dissipate any deviations 37 1C). From the laws of thermodynamics, when a from the equilibrium condition, and short circuit the þ solution is already at chemical equilibrium with need for CA activity. Only very large fluxes of H ,CO2 À respect to these species, CA activity cannot have an or HCO3 could overload diffusive dissipation. The effect. To provide substrate for CA activity, the physiological need for CA activity is therefore set by concentration of at least one of the three species must the balance between the magnitude of fluxes that be changed at a rate that exceeds the spontaneous re- disturb equilibrium and the capacity for their diffusive equilibration rate of the buffer. Figure 2c illustrates an dissipation. experiment in which the equilibrium between extracel- Diffusive restrictions are predicted to raise the þ À lular H ,CO2 and HCO3 was disturbed, providing importance of CA activity. To test this, experiments substrate for CAIX catalysis. The pH at the extra- were performed on multi-cellular spheroids that harbour cellular surface of ca9-transfected cells of the HCT116 a diffusionally restricted space surrounding component line was probed using a glycocalyx-binding pH dye cells (Sutherland, 1988; Swietach et al., 2008; Swietach (fluorescein–wheat germ agglutinin conjugate; Stock et al., 2009). Figure 3 illustrates an experiment per- et al., 2007). Cells were bathed in pre-equilibrated 5% formed on a cultured spheroid, comprised of CAIX À CO2/12mM HCO3 buffer at pH 7.2 and subjected to a over-expressing HCT116 cells, superfused with CO2/ À brief ‘ammonium prepulse’ solution manoeuvre. Rapid HCO3 buffer. The spheroid was first subjected to an exposure to solution containing ammonium chloride ‘ammonium prepulse’ (Leem et al., 1999; Swietach et al., (NH4Cl) produced a rapid transmembrane influx of 2008). During this solution manoeuvre, a rapid intra- þ NH3, and drives extracellular NH4 deprotonation. The cellular alkalinisation is followed by a more gradual þ release of H ions at the extracellular surface of the cell recovery of pHi. On removal of ammonium from the produces a measureable surface pHe transient. Removal superfusate, an intracellular acid load is induced, so that þ of NH4Cl drives these reactions in the reverse direction, pHi becomes lower than control levels. This H load is leading to an opposite surface pHe response. Although subsequently removed by acid extruders. Recovery of addition and subsequent withdrawal of ammonium- pHi at the periphery of the spheroid was found to be containing solution alkalinises and then acidifies the considerably faster than at the core (Figure 3a). This can intracellular compartment, only extracellular pH be explained in terms of diffusion distances in the changes are registered by the glycocalyx-binding dye. extracellular space within the spheroid, which must be The size of the surface pHe transient depends on the crossed to complete the process of acid removal. À À ability of extracellular CO2/HCO3 to buffer pHe Extracellular CO2/HCO3 has a dual role in this process. À changes (Swietach et al., 2009). When the experiment First, CO2/HCO3 functions as a mobile buffer for was repeated in the presence of the CA inhibitor shuttling H þ ions across the restricted extracellular acetazolamide, surface pHe transients were much larger space of the spheroid (Figure 3ci). Second, it is a source À than under control conditions. Under this experimental of HCO3 for cellular uptake (Figure 3cii). The protocol, CAIX activity reduces out-of-equilibrium subsequent intracellular reaction between H þ ions and À À episodes of CO2/HCO3 buffer. sequestered HCO3 ions removes excess cellular acid. þ À þ The driving force that disturbs H /CO2/HCO3 CAs can have a role in facilitating H diffusion, in À equilibrium in Figure 2c (rapid transmembrane NH3 supplying extracellular HCO3 for membrane transport, À þ flux) is not a common physiological phenomenon. More and in catalysing intracellular HCO3 þ H titration. To À þ usual perturbations would include HCO3 and H test the impact of CA inhibition on pH regulation in our membrane transport or transmembrane CO2 flux. spheroid model, experiments were repeated in the Above a certain critical magnitude, these fluxes may presence of acetazolamide. CA inhibition had little provide enough substrate for measureable CAIX effects. effect on the pHi recovery time course at the spheroid The coupling between membrane flux and CA activity periphery, but slowed recovery significantly at the core has led to the proposal of an obligatory physical binding (Figure 3b, see also inset pHi maps). The importance of between CA and transporters, called a transport CA catalysis increases with distance, as diffusive

Oncogene Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6513

Figure 3 CA activity facilitates acid-extrusion in spheroid models. The rate of acid extrusion was measured confocally in spheroids (radius 110±5 mm) made of CAIX over-expressing HCT116 cells, loaded with the intracellular pH reporter dye, carboxy-SNARF-1 (average of six spheroids). (a) Spheroids were subjected to an ammonium prepulse solution manoeuvre, which produces an intracellular acid load. This acid load is then removed across cell membranes (by transporters) and across the extracellular domain within the spheroid (by diffusion) to the surrounding superfusate. Recovery of pHi was markedly faster in the spheroid periphery (outer 11 mm ring) than in deeper layers (inner 44 mm core), in agreement with diffusion distances. (b) The experiment was repeated in the presence of 100 mM acetazolamide (ATZ) to block all CA activity, including CAIX. Recovery of pHi was not greatly affected at the periphery, but significantly slowed in the core. This gave rise to pronounced pHi non-uniformity during acid removal. Insets show pHi maps after þ 10 min of pHi recovery in layered regions of interest, 11 mm thick each. (c) CAIX expressed in these spheroids may (i) facilitate H diffusion or (ii) facilitate H þ extrusion.

coupling between acid-producing cells and the bulk CAIX can influence steady state intracellular and À superfusate becomes weaker, making CO2/HCO3 buffer extracellular pH more prone to dysequilibrium. In summary, a role for CAIX emerges in diffusionally When membrane transport is rate-limited by the restricted spaces, wherein diffusive coupling with the diffusion of its substrate, the overall flux is described superfusate (or blood) is no longer a viable means of as a diffusion-reaction phenomenon. The combination þ À keeping H /CO2/HCO3 at equilibrium. Under these of sustained membrane transport and restricted sub- conditions, catalysis by CAIX offers an alternative strate diffusion can produce a steady state in which the þ À means of maintaining H /CO2/HCO3 equilibrium. It is participating solutes are not in equilibrium. One of interest to reiterate that the ca9 gene is hypoxia example of this is mitochondrial respiration. CO2 sensitive (Wykoff et al., 2000; Ivanov et al., 2001). As produced continuously by mitochondria diffuses down oxygen gradients are a physiological indicator of its concentration gradient towards blood (or super- diffusion distance, CAIX will be expressed naturally in fusate). In parallel to this diffusive event, CO2 can regions of tissue in which demand for its catalytic undergo hydration en route. Out-of-equilibrium concen- activity is greatest. tration gradients of CO2 and its hydration products

Oncogene Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6514 À þ (HCO3 and H ) can arise, particularly when the phenylhydrazone in glucose-containing media reduced þ À diffusional driving forces for H , HCO3 or CO2 are the effect of CAIX on pHe. However, equilibration of not equal. This model predicts that respiring tissues spheroids with a-cyano-4-hydroxycinnamate, an inhibi- þ should develop spatial gradients of pHi and pHe that are tor of lactic acid (H -monocarboxylate) transporters, sensitive to CA activity. Moreover, these pH gradients did not change the ability of CAIX to reduce pHe. will also depend on the site of CA activity (Swietach Overall, these observations suggest that aerobic respira- et al., 2008; Swietach et al., 2009). For instance, a tion is a major source of CO2 for CAIX activity. dominance of intracellular CA activity would favour Spheroids as large as 500 mm in radius were still able to intracellular CO2 hydration, which lowers pHi. produce substrate for CAIX (Figure 4c), even though The role of CAIX in shaping pHi/pHe gradients was mitochondria are expected to be far less effective at studied in spheroids grown from CAIX over-expressing respiring under the hypoxic conditions that persist cells (RT112 renal bladder or HCT116 colon cancer beyond a depth of 100–200 mm. As explained earlier lines). Figure 4 shows experimental data from spheroid (Figure 1), CO2 may be produced in anaerobic regions À models at steady state, superfused with physiological by a titration reaction between lactic acid and HCO3. À CO2/HCO3 buffer (Swietach et al., 2008; Swietach et al., Alternatively, CAIX at the anaerobic core may catalyse 2009). A major acid product of these spheroids is CO2, the hydration of CO2 produced remotely in more produced in aerobic regions (up to a depth of 100– peripheral, aerobic cells. Indeed, there is no necessity 200 mm) by mitochondria, and in anaerobic regions for the substrate for CAIX to be produced by the CAIX- elsewhere from the titration reaction between acid and expressing cell. À HCO3. Because of the persistent CO2 production and The physiological role of CAIX is difficult to assess continuous superfusion, spheroids will develop spatial without a complete diffusion-reaction description of gradients of pHi (Figure 4a) and pHe (Figure 4b). Intra- acid/base fluxes. To study CAIX function quantita- or extracellular pH was imaged confocally with a cell tively, a mathematical model is required to test CAIX loaded pH dye (carboxy-SNARF-1) or a membrane- catalysis in the context of realistic physiological H þ , À impermeant dye (fluorescein-5-(and-6-)-sulphonic acid), CO2 and HCO3 fluxes. Figure 5 shows the results of respectively. The effect of CAIX activity on pH was such a model, on the basis of the parameterisations assessed from the effect of membrane impermeant, small obtained using published work (Swietach et al., 2009). molecule inhibitors that target extracellular CA activity, The model was produced for a spheroid consisting of an that is, principally CAIX in the experimental model intracellular space (80% of spheroid volume) and a used. Figures 4a and 4bi show the radial profiles of pHi restricted extracellular space that is coupled diffusively and pHe gradients, respectively, measured from the edge with the bulk superfusate. The model was run for two to the core of spheroids. CAIX activity reduced the rates of glucose metabolism, representing lower magnitude of intracellular pH gradients, making the (0.5 mmoles/l/min; Vaupel et al., 1989) and upper cells of the spheroid-core less acidic (Figure 4a). In (5 mmoles/l/min; Kunz-Schughart et al., 2000) estimates contrast, CAIX activity increased the magnitude of for cancer cell respiratory rates. Three scenarios were extracellular pH gradients, making the milieu surround- modelled, each for a range of spheroid radii. In the first ing cells at the spheroid core more acidic. Catalysis by model (the ‘mitochondrial model’; Figure 5a), all þ CAIX shifts the site of H -yielding CO2 hydration from metabolised glucose was assumed to be converted to an intracellular to an extracellular locus, and thereby 6 Â CO2 by aerobic respiration. In models two and facilitates the removal of cellular acid. The amount of three, each glucose molecule was assumed to be acid deposited in the extracellular domain of the metabolised to two lactic acid molecules. For model À spheroid core increased with spheroid radius two (‘HCO3 titration’), lactic acid was neutralised by À (Figure 4bii), as expected from the increase in diffusion means of reaction with HCO3 (Figure 1b). In this distance. model, the cell emits acid across its membrane in the þ The acidifying effect of CAIX activity in the extra- form of CO2. In the third model (‘H extrusion’), lactic cellular domain of the spheroid confirms that the acid was extruded directly across the membrane enzyme is performing net catalysis in the direction of (Figure 1a). CO2 hydration. The source of CO2 was studied further The model outputs plotted in Figure 5 show the by altering cellular metabolism with drugs and meta- steady state pHi or pHe predicted for the core of bolic substrates in CAIX over-expressing or sham- spheroids, as a function of its radius. Arrows at the top transfected (‘empty vector’) spheroids (Figure 4c) left corner of each plot depict the direction of pH change (Swietach et al., 2008; Swietach et al., 2009). Replacing brought about by CAIX catalysis. CAIX activity medium glucose with its non-metabolisable derivative, increased pHi under conditions in which cells emit acid 2-deoxyglucose, reduced overall metabolic rate and, as in the form of CO2, that is, in the ‘mitochondrial’ expected, greatly reduced extracellular acidification. (Figure 5a) and ‘titration’ (Figure 5b) models. Of the Incubation with galactose in place of glucose favours two models, the former imposes a greater acid load aerobic respiration because the glycolysis of galactose within a cell because of its higher yield of acid does not yield energy. CAIX-dependent extracellular equivalents (6 Â CO2 versus 2 Â CO2 per glucose). The acidification was still observed under these conditions. CAIX-dependent alkalinisation of the intracellular Inhibition of mitochondrial respiration with rotenone, space was accompanied by a reduction in pHe. These myxothiazol or carbonyl cyanide p-(trifluoromethoxy) effects were more prominent at the higher respiratory

Oncogene Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6515

Figure 4 CAIX activity shapes steady state intra- and extracellular pH gradients in spheroid models. (a) Steady state intracellular pH was measured confocally with carboxy-SNARF-1 in spheroids made of (i) RT112 or (ii) HCT116 cells over-expressing CAIX. Inhibition of CAIX with a membrane-impermeant inhibitor increased the size of pHi gradients (blue shading between curves denotes alkalinising effect of CAIX), as seen in the insets for a sample spheroid. (b) Steady state extracellular pH was measured confocally with fluorescein-5-(and-6-)-sulphonic acid in spheroids made of HCT116 cells over-expressing CAIX. (i) Inhibition of CAIX with membrane-impermeant drugs reduced pHe gradients (red shading between curves denotes acidifying effect of CAIX). (ii) Extracellular acidity increased with spheroid radius. At any given spheroid radius, CAIX activity decreases pHe further, suggesting that the enzyme catalyses in the direction of CO2 hydration. (c)pHe at the core of spheroids made of ca9 transfected (black) or sham transfected (grey) HCT116 cells measured under a variety of conditions. The ability of CAIX to reduce pHe was blocked by inhibiting mitochondria with rotenone (Rot), myxothiazol (Myxo) or carbonyl cyanide p-(trifluoromethoxy) phenylhydrazone (FCCP), but not affected by blocking lactic acid efflux with a-cyano-4-hydroxycinnamate (Cinn). Replacing medium glucose with galactose (which suppresses anaerobic respiration) had no effect on CAIX-dependent extracellular acidification. Replacing glucose with 2-deoxyglucose to decrease respiratory rate reduced extracellular acidification. Data from Swietach et al. (2008) and Swietach et al. (2009). * Denotes significant # difference (Po0.05) between empty vector and CAIX-expressor spheroid. Denotes significant change in pHe gradient between the control and test CAIX-expressing spheroid.

Oncogene Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6516

Figure 5 Mathematical model of the effect of CAIX activity on steady state pH. A diffusion-reaction model, based on Swietach et al. (2009), was run to simulate the effect of CAIX (100-fold catalysis) on spheroid pHi and pHe as a function of spheroid radius. Glucose respiration was set to either 5.0 or 0.5 mM/min (the upper and lower estimates for cancer cell metabolism). Intra- and extracellular À intrinsic (non-CO2/HCO3) buffering was set to 20 and 3 mM, respectively. In all, 80% of the spheroid mass was intracellular space. Restricted diffusion was allowed in the extracellular domain, in communication with the bulk superfusate (maintained at pH 7.4, À buffered by 5% CO2/22 mM HCO3). (a) Glucose is respired to 6 Â CO2 by mitochondria. (b) Glucose is respired to 2 Â lactic acid. Acid À equivalents are then excreted from cells in the form of CO2, produced from the intracellular titration reaction with intracellular HCO3, taken up by membrane transport. (c) Glucose is respired to 2 Â lactic acid. H þ ions are then extruded across the membrane. Blue and red shading denote an alkalinising and acidifying effect of CAIX activity, respectively.

þ À þ À rate, as this tends to disturb H /CO2/HCO3 equilibrium is more likely to disturb H /CO2/HCO3 equilibrium. À more profoundly. The ability of CAIX to affect steady This is expected, given that HCO3 titration involves two À state pHi varied with diffusion distance, and was most opposite fluxes (CO2 out, HCO3 in), each on its own þ À pronounced for spheroid radii of between B50 and disturbing H /CO2/HCO3 equilibrium, as opposed to B250 mm, that is, diffusion distances that overlap with just one flux (CO2 out) under the mitochondrial model. the onset of hypoxia and the activation of by hypoxia- Under the H þ extrusion model, net CAIX catalysis is inducible-factor-sensitive pathways. This arrangement is in the direction of extracellular CO2 production rather intriguing in the context of in vivo CAIX expression in than CO2 hydration. Under these circumstances, CAIX tumours, which normally show CAIX staining in a activity is predicted to reduce pHi and raise pHe. The þ hypoxic rim of tissue just beyond the better-perfused spontaneous rate of the reaction between HCO3 and H aerobic layer of cells (Beasley et al., 2001; Airley et al., is several orders of magnitude higher than CO2 2003; Figure 6a). The hypoxia response element for the hydration, therefore the capacity for CAIX to catalyse 0 ca9 gene overlaps with the 5 start site for transcription, HCO3 protonation is limited. For this reason, the making CAIX one of the most sensitive and responsive quantitative effects of CAIX under the third model are proteins to hypoxic regulation (Figure 6b). This may considerably smaller than under the first two models. suggest that the evolution of CAIX regulation is closely Using the mathematical predictions, it becomes associated with its function (Wykoff et al., 2000). At possible to assess the direction of net CAIX activity depths less than B50 mm, CAIX activity becomes from the effect of CAIX inhibitors on pHi and/or pHe redundant, as diffusion distances are too short to gradients. The experimental data and simulations þ À perturb H /CO2/HCO3 equilibrium. The model also presented thus far show that CAIX catalysis, under reveals that CAIX was able to alter steady state pHi/pHe appropriate conditions, shapes pHi and pHe gradients in more profoundly and at a greater depth under the tissues. This predicts an important role for extracellular À HCO3 titration model, suggesting that this arrangement CA isoforms in overall cell biology of inadequately

Oncogene Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6517 1984; , 1986; Matsuyama et al., 1985, 2000). Proof-of- principle experiments have supported a correlation between pH regulation and cell survival (Pouyssegur et al., 1984), and it has been proposed that cells with the most powerful pH regulatory mechanisms should have the best survival prospects. The ability of tumour cells to out-compete normal host cells in a Darwinian battle for resources and space has drawn attention to pH regulation in cancer. The maintenance of healthy acid/ base balance in solid tumours is challenged by their high metabolic rate (which tends to increase acid loading of cells) and inadequate blood perfusion (which slows the removal of acid from cells). Nonetheless, nuclear magnetic resonance studies have shown that solid tumours are successful in maintaining an alkaline pHi (Gillies et al., 2004). The acid that is so potently removed from cells lowers pHe below 7.0 (Lindner and Raghavan, 2009) and contributes to one of the most striking hallmarks of cancer, an acidic extracellular milieu (Griffiths et al., 2001). Low pHe may favour invasive behaviour by degrading the extracellular matrix (Martinez-Zaguilan et al., 1996; Giusti et al., 2008), triggering signalling cascades for proliferation via G-protein coupled surface H þ receptors (Huang et al., 2008), and exerting selection pressure that favours cancer survival (Fang et al., 2008) and suppresses natural immunity (Lardner, 2001). From a therapeutic point of view, low pH can confer a degree of drug Figure 6 Diffusion distance versus CAIX expression. Figure from e Beasley et al. (2001). (a) CAIX immunostaining in sections of head resistance to tumours because of the tendency for and neck squamous cell carcinoma, showing range of distances weakly basic drugs to become protonated and imper- between a central blood vessel (labelled with CD34 monoclonal meant at acidic pH (Sauvant et al., 2008). antibody) and the start of CAIX expression (M75 antibody) and The combination of low pH and high pH in tumours necrosis. The median distance from the blood vessel to the start of e i necrosis was 130 mm, and from the blood vessel to the start must almost certainly be achieved by an elevated of CAIX expression was 80 mm. CAIX expression is associated with capacity for transmembrane acid extrusion. The most a hypoxic yet viable rim of tissue. Magnification  250. studied candidates for this activity are membrane (b) Prediction for the longitudinal O2 gradient from a blood vessel proteins such as H þ pumps (Martinez-Zaguilan et al., (at distance ¼ 0 mm), based on a formula derived by Thomlinson 1993), Na þ /H þ exchange (Sardet et al., 1989; Counillon and Gray (1955), derived for squamous cell carcinoma of the lung. þ CAIX expression (shaded) is predicted to coincide with hypoxia and Pouyssegur, 2000), H -lactate co-transport À below 1% O2. (Halestrap and Meredith, 2004) and various HCO3 trans- porters (Romero et al., 2004; Alper, 2006). A recent addition to this list is CAIX. Unlike active transporters, perfused tissues such as solid tumours or normal tissue CAIX catalysis does not put pressure on the already under periods of ischaemia. Tissues in vivo are likely to stretched cellular energy supplies. Instead, CAIX can involve aerobic respiration in peripheral, well-oxyge- produce the desired effect (lowering pHe and raising nated regions (Figure 5a) and anaerobic respiration at pHi) by favouring CO2 hydration in the extracellular greater depths. Metabolic acid in anaerobic regions will domain (Figure 4). The only requirement for this À þ undergo HCO3 titration (Figure 5b) or H extrusion outcome is for most of the transmembrane acid-traffic (Figure 5c), depending on the expression and activity of to be carried in the form of CO2. It is expected that membrane-bound transporters. The effect of CAIX on CAXII, another hypoxia-induced extracellular isoform steady state pHi and pHe can be complex and should be of CA, should show similar physiology to CAIX. analysed in the context of depth (that is, oxygenation) Over the past three decades, the most intensively and membrane transport. studied acid extruder has been the NHE. Mutant cell lines lacking NHE isoform 1 (NHE1) were less likely to form tumours in mouse models (Lagarde et al., 1988) as a result of weakened pHi regulation (Pouyssegur et al., CAIX may have an important role in tissue growth 1984). The role of NHE1 in supporting tumour growth and development under different regimes of acid loading (lactic acid versus CO2) was studied in mutant cells lacking the Many cellular processes, such as metabolism, prolifera- capacity to respire aerobically or anaerobically (Franchi tion, growth and migration, are highly pH sensitive et al., 1981). In the absence of aerobic metabolism (resÀ (Wohlhueter and Plagemann, 1981; Pouyssegur et al., mutants), tumour growth was greatly reduced, suggesting

Oncogene Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6518 that mitochondrial respiration is important for growth pHe that favours growth. One example of such a and development. Mutating NHE1 to an inactive form mechanism could be the expression of V-type H þ in resÀ cells reduced growth further, in support of an ATPase at the plasmalemma (Martinez-Zaguilan et al., important role for NHE in removing the large load of 1993). It has been shown that in aggressive cell lines, lactic acid produced by resÀ cells. However, glycolysis- plasmalemmal expression of this H þ pump is high deficient cells (glyÀ mutants) were still able to produce (Sennoune et al., 2004) because of the elevated net tumours in mice, suggesting that lactic acid production trafficking from the endosomal pool of V-type ATPases is not a phenotype necessary for tumour growth. (Martinez-Zaguilan et al., 1999). Plasmalemmal activity Inactivating NHE1 in glyÀ mutants had no significant of this pump would produce an acidic extracellular effect, indicating that under circumstances in which environment and alkalinise the cytoplasm, akin to the lactic acid is not the major acid product, NHE1 is not effect of expressing extracellular CAs in CO2-producing essential for supporting growth and development. tumours. Indeed, subsequent studies with glycolysis-deficient cells The induction of CAIX and its tight regulation by have proposed that lactic acid is not the only source of hypoxia in many tumours suggest that elevated CO2 extracellular acidity in solid tumours (Newell et al., production may be a hallmark of tumour physiology. A 1993). Recent measurements in colonic tumours have significant fraction of CO2 will come from oxidative shown that CO2 is the principal form in which acid is metabolism in the mitochondria, which will be function- removed from cells (Holm et al., 1995; Griffiths et al., ing at a sub-optimal level in hypoxia, yet still generating 2001). These findings have suggested that in CO2- ATP at o1% O2. However, other metabolic processes emitting tumours, transporters other than NHE, such essential for tumour growth also generate CO2,suchas À as HCO3 carriers (L’Allemain et al., 1985; Lee and the pentose shunt and glutaminolysis. It will be of Tannock, 1998), may have a major part in regulating interest to investigate the importance of CAIX for pHi, and supporting growth and development. The tumours that require these pathways predominately for À emerging importance of HCO3 transport and CO2 growth. An unusual feature for a hypoxia-induced venting has led to speculation about the role of protein is the prolonged half-life (2–3 days) of the extracellular CA isoforms in tumour growth and induced enzyme, compared to the half-life of its RNA. development. This may explain discrepancies between expression of The effect of CAIX on tumour growth has been CAIX and other biochemical markers of hypoxia, such studied in cells using gene silencing (Chiche et al., 2009). as pimonidazole (Wykoff et al., 2000; Turner et al., Knock down of ca9 reduced spheroid growth in vitro 2002). There may thus be role for CAIX in regulating and xenograft growth in nude mice by 40%. An pH under specific circumstances of recovery from additional silencing of ca12 reduced xenograft growth hypoxia, which is accompanied by a large CO2 load. by a further 45%. Growth of tumour xenografts was Such re-oxygenated tissues may require CAIX for the also retarded by albumin-binding acetazolamide deriva- recovery period and this aspect of CAIX physiology tives that target extracellular CA isoforms (Ahlskog requires further study. et al., 2009). These dramatic effects illustrate the importance of extracellular CA activity and may explain the correlation between CAIX/CAXII expression in tumour tissue samples and poor patient prognosis Concluding remarks (Saarnio et al., 1998; Liao and Stanbridge, 2000; Chia et al., 2001; Koukourakis et al., 2001; Loncaster et al., It is perhaps not surprising that a process as funda- 2001; Potter and Harris, 2003; Haapasalo et al., 2006; mental as pH regulation is performed by a number of Haapasalo et al., 2008). These results make a case for seemingly redundant mechanisms. Failure of one system clinical trials for studying the responsiveness of tumours could be compensated by another. Despite the hetero- in patients to anti-CAIX/anti-CAXII therapies based on geneity of pH regulatory mechanisms, all transport small molecule inhibitors (Supuran, 2008; Guler et al., strategies fall into one of the two categories illustrated in 2010) or selective antibodies with inhibitory effects (Xu Figure 1. Both schemes result in acid removal, but have þ À et al., 2010). It is noteworthy, however, that some different substrate requirements (H versus HCO3) and studies on gastric and oesophageal tumours have shown may be subject to contrasting control mechanisms. This a lack of positive correlation between CAIX expression highlights the importance of careful experimental de- À and cancer progression (Pastorekova et al., 1997; Turner sign. For example, HCO3 transporters produce physio- et al., 1997). This may reflect different basal effects, in logical acid/base fluxes only when they are studied À which CAIX is expressed in normal tissue but may be in cells superfused with CO2/HCO3-buffered media down-regulated because of the loss of differentiation. In (Thomas, 1989). To study pH regulation with relevance addition, there is no standardised scoring system for to solid tumours, it is necessary to mimic the extra- positivity between different studies and the antibodies cellular composition with respect to CO2 partial À used. Nevertheless, these findings question the general pressure, HCO3 concentration and pH. This is also true prognostic value of CAIX and the suitability of CAIX- for studies involving CAIX in isolated cells. In the À directed therapy in certain types of cancer. In aggressive nominal absence of superfusate CO2/HCO3, CAIX À cancer without up-regulated CAIX/CAXII, other me- activity may become crucial for supplying HCO3 chanisms may operate to produce a tumour pHi and transporters with substrate. Under more physiological

Oncogene Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6519 À extracellular CO2 and HCO3 levels, and with good homeostasis. The CAIX cytoplasmic tail, with its superfusate-cell diffusive coupling, such a role for CAIX multiple phosphorylation sites, may regulate interac- catalysis may be of lesser importance. Assessments of the tions between CAIX and other proteins (Dorai et al., role of CAIX must be made in the context of the 2005; Hulikova et al., 2009). The proteoglycan domain experimental protocol performed. of CAIX (Figure 1aii), that is not present in CAXII, It is noteworthy that in the discussion so far, the may also serve a physiological role that is independent effects of CAIX expression have been described in terms of CA activity. Such catalysis-independent effects can be of the enzyme’s CA activity. It is plausible that the distinguished experimentally by their insensitivity to CA protein itself may interact with other cellular elements inhibitors such as acetazolamide. and produce effects that are independent of its catalytic domain. These interactions may, for instance, involve coordinated expression. We and others have shown that, at least in some cell lines, CAIX over-expression alters Conflict of interest the expression pattern of other CA isoforms, including CAII (Pan et al., 2006; Swietach et al., 2008) and CAXII The authors declare no conflict of interest. (Chiche et al., 2009). The interaction and co-regulation of CA isoforms is a developing area, with important implications for effectiveness of therapies targeting Acknowledgements extracellular-facing CAs. It is possible that CAIX expression may also alter the expression or turnover This study was supported by the Royal Society, Medical (Morgan et al., 2007) of membrane acid/base transpor- Research Council, British Heart Foundation, the Cancer ters, and thereby exert a more complex effect on pHi Research UK and the European Union Framework 7 Metoxia.

References

Ahlskog JK, Dumelin CE, Trussel S, Marlind J, Neri D. (2009). In vivo tumor cell growth by counteracting acidosis through the regulation targeting of tumor-associated carbonic anhydrases using acetazola- of the intracellular pH. Cancer Res 69: 358–368. mide derivatives. Bioorg Med Chem Lett 19: 4851–4856. Counillon L, Pouyssegur J. (2000). The expanding family of eucaryotic Airley RE, Loncaster J, Raleigh JA, Harris AL, Davidson SE, Hunter Na(+)/H(+) exchangers. J Biol Chem 275: 1–4. RD et al. (2003). GLUT-1 and CAIX as intrinsic markers of De Simone G, Supuran CT. (2010). Carbonic anhydrase IX: hypoxia in carcinoma of the cervix: relationship to pimonidazole biochemical and crystallographic characterization of a novel binding. Int J Cancer 104: 85–91. antitumor target. Biochim Biophys Acta 1804: 404–409. Alper SL. (2006). Molecular physiology of SLC4 anion exchangers. Dorai T, Sawczuk IS, Pastorek J, Wiernik PH, Dutcher JP. (2005). The Exp Physiol 91: 153–161. role of carbonic anhydrase IX overexpression in kidney cancer. Eur Alterio V, Hilvo M, Di Fiore A, Supuran CT, Pan P, Parkkila S et al. J Cancer 41: 2935–2947. (2009). Crystal structure of the catalytic domain of the tumor- Fang JS, Gillies RD, Gatenby RA. (2008). Adaptation to hypoxia and associated human carbonic anhydrase IX. Proc Natl Acad Sci USA acidosis in carcinogenesis and tumor progression. Semin Cancer Biol 106: 16233–16238. 18: 330–337. Beasley NJ, Wykoff CC, Watson PH, Leek R, Turley H, Gatter K Franchi A, Silvestre P, Pouyssegur J. (1981). A genetic approach to the et al. (2001). Carbonic anhydrase IX, an endogenous hypoxia role of energy metabolism in the growth of tumor cells: tumor- marker, expression in head and neck squamous cell carcinoma and igenicity of fibroblast mutants deficient either in glycolysis or in its relationship to hypoxia, necrosis, and microvessel density. Cancer respiration. Int J Cancer 27: 819–827. Res 61: 5262–5267. Gatenby RA, Gillies RJ. (2004). Why do cancers have high aerobic Boron WF. (2004). Regulation of intracellular pH. Adv Physiol Educ glycolysis? Nat Rev Cancer 4: 891–899. 28: 160–179. Generali D, Fox SB, Berruti A, Brizzi MP, Campo L, Bonardi S et al. Boron WF. (2010). Evaluating the role of carbonic anhydrases in the (2006). Role of carbonic anhydrase IX expression in prediction of transport of HCO3–related species. Biochim Biophys Acta 1804: the efficacy and outcome of primary epirubicin/tamoxifen therapy 410–421. for breast cancer. Endocr Relat Cancer 13: 921–930. Chambard JC, Pouyssegur J. (1986). Intracellular pH controls growth Giatromanolaki A, Koukourakis MI, Sivridis E, Pastorek J, Wykoff factor-induced ribosomal protein S6 phosphorylation and protein CC, Gatter KC et al. (2001). Expression of hypoxia-inducible synthesis in the G0—G1 transition of fibroblasts. Exp Cell Res 164: carbonic anhydrase-9 relates to angiogenic pathways and indepen- 282–294. dently to poor outcome in non-small cell lung cancer. Cancer Res Chegwidden WR, Dodgson SJ, Spencer IM. (2000). The roles of 61: 7992–7998. carbonic anhydrase in metabolism, cell growth and cancer in Gillies RJ, Martinez-Zaguilan R, Peterson EP, Perona R. (1992). Role animals. Exs 90: 343–363. of intracellular pH in mammalian cell proliferation. Cell Physiol Chia SK, Wykoff CC, Watson PH, Han C, Leek RD, Pastorek J et al. Biochem 2: 159–179. (2001). Prognostic significance of a novel hypoxia-regulated marker, Gillies RJ, Martinez-Zaguilan R, Martinez GM, Serrano R, Perona R. carbonic anhydrase IX, in invasive breast carcinoma. J Clin Oncol (1990). Tumorigenic 3T3 cells maintain an alkaline intracellular pH 19: 3660–3668. under physiological conditions. Proc Natl Acad Sci USA 87: 7414–7418. Chiche J, Brahimi-Horn MC, Pouysse´gur J. (2010). Tumour hypoxia Gillies RJ, Raghunand N, Garcia-Martin ML, Gatenby RA. (2004). induces a metabolic shift causing acidosis: a common feature in pH imaging. A review of pH measurement methods and applica- cancer. J Cell Mol Med 14: 771–794. tions in cancers. IEEE Eng Med Biol Mag 23: 57–64. Chiche J, Ilc K, Laferriere J, Trottier E, Dayan F, Mazure NM et al. Giusti I, D’Ascenzo S, Millimaggi D, Taraboletti G, Carta G, (2009). Hypoxia-inducible carbonic anhydrase IX and XII promote Franceschini N et al. (2008). Cathepsin B mediates the pH-dependent

Oncogene Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6520 proinvasive activity of tumor-shed microvesicles. Neoplasia 10: Leem CH, Lagadic-Gossmann D, Vaughan-Jones RD. (1999). 481–488. Characterization of intracellular pH regulation in the guinea-pig Gleadle JM, Ratcliffe PJ. (1998). Hypoxia and the regulation of gene ventricular myocyte. J Physiol 517(Part 1): 159–180. expression. Mol Med Today 4: 122–129. Liao SY, Stanbridge EJ. (2000). Expression of MN/CA9 protein in Griffiths JR, McIntyre DJ, Howe FA, Stubbs M. (2001). Why are Papanicolaou smears containing atypical glandular cells of un- cancers acidic? A carrier-mediated diffusion model for H+ transport determined significance is a diagnostic biomarker of cervical in the interstitial fluid. Novartis Found Symp 240: 46–62. dysplasia and neoplasia. Cancer 88: 1108–1121. Griffiths JR, Stevens AN, Iles RA, Gordon RE, Shaw D. (1981). 31P- Lindner D, Raghavan D. (2009). Intra-tumoural extra-cellular pH: a NMR investigation of solid tumours in the living rat. Biosci Rep 1: useful parameter of response to chemotherapy in syngeneic tumour 319–325. lines. Br J Cancer 100: 1287–1291. Guler OO, De Simone G, Supuran CT. (2010). Drug design studies of Loncaster JA, Harris AL, Davidson SE, Logue JP, Hunter RD, the novel antitumor targets carbonic anhydrase IX and XII. Curr Wycoff CC et al. (2001). Carbonic anhydrase (CA IX) expression, a Med Chem 17: 1516–1526. potential new intrinsic marker of hypoxia: correlations with tumor Haapasalo J, Hilvo M, Nordfors K, Haapasalo H, Parkkila S, oxygen measurements and prognosis in locally advanced carcinoma Hyrskyluoto A et al. (2008). Identification of an alternatively spliced of the cervix. Cancer Res 61: 6394–6399. isoform of carbonic anhydrase XII in diffusely infiltrating astrocytic Martinez-Zaguilan R, Lynch RM, Martinez GM, Gillies RJ. (1993). gliomas. Neuro Oncol 10: 131–138. Vacuolar-type H(+)-ATPases are functionally expressed in plasma Haapasalo JA, Nordfors KM, Hilvo M, Rantala IJ, Soini Y, Parkkila membranes of human tumor cells. Am J Physiol 265: C1015–C1029. AK et al. (2006). Expression of carbonic anhydrase IX in astrocytic Martinez-Zaguilan R, Raghunand N, Lynch RM, Bellamy W, tumors predicts poor prognosis. Clin Cancer Res 12: 473–477. Martinez GM, Rojas B et al. (1999). pH and drug resistance. I. Halestrap AP, Meredith D. (2004). The SLC16 gene family-from Functional expression of plasmalemmal V-type H+-ATPase in monocarboxylate transporters (MCTs) to aromatic amino acid drug-resistant human breast carcinoma cell lines. Biochem Pharma- transporters and beyond. Pflugers Arch 447: 619–628. col 57: 1037–1046. Harris AL. (2002). Hypoxia–a key regulatory factor in tumour growth. Martinez-Zaguilan R, Seftor EA, Seftor RE, Chu YW, Gillies RJ, Nat Rev Cancer 2: 38–47. Hendrix MJ. (1996). Acidic pH enhances the invasive behavior of Holm E, Hagmuller E, Staedt U, Schlickeiser G, Gunther HJ, human melanoma cells. Clin Exp Metastasis 14: 176–186. Leweling H et al. (1995). Substrate balances across colonic Matsuyama S, Llopis J, Deveraux QL, Tsien RY, Reed JC. (2000). carcinomas in humans. Cancer Res 55: 1373–1378. Changes in intramitochondrial and cytosolic pH: early events Huang WC, Swietach P, Vaughan-Jones RD, Ansorge O, Glitsch MD. that modulate caspase activation during apoptosis. Nat Cell Biol 2: (2008). Extracellular acidification elicits spatially and temporally 318–325. distinct Ca2+ signals. Curr Biol 18: 781–785. McLean LA, Roscoe J, Jorgensen NK, Gorin FA, Cala PM. (2000). Hulikova A, Zatovicova M, Svastova E, Ditte P, Brasseur R, Malignant gliomas display altered pH regulation by NHE1 Kettmann R et al. (2009). Intact intracellular tail is critical for compared with nontransformed astrocytes. Am J Physiol Cell proper functioning of the tumor-associated, hypoxia-regulated Physiol 278: C676–C688. carbonic anhydrase IX. FEBS Lett 583: 3563–3568. Morgan PE, Pastorekova S, Stuart-Tilley AK, Alper SL, Casey JR. Innocenti A, Pastorekova S, Pastorek J, Scozzafava A, De Simone G, (2007). Interactions of transmembrane carbonic anhydrase, CAIX, with Supuran CT. (2009). The proteoglycan region of the tumor- bicarbonate transporters. Am J Physiol Cell Physiol 293: C738–C748. associated carbonic anhydrase isoform IX acts as anintrinsic buffer Nakhoul NL, Davis BA, Romero MF, Boron WF. (1998). Effect of optimizing CO2 hydration at acidic pH values characteristic of solid expressing the water channel aquaporin-1 on the CO2 permeability tumors. Bioorg Med Chem Lett 19: 5825–5828. of Xenopus oocytes. J Physiol 274: C543–C548. Irving M, Maylie J, Sizto NL, Chandler WK. (1990). Intracellular Newell K, Franchi A, Pouyssegur J, Tannock I. (1993). Studies with diffusion in the presence of mobile buffers. Application to proton glycolysis-deficient cells suggest that production of lactic acid is movement in muscle. Biophys J 57: 717–721. not the only cause of tumor acidity. Proc Natl Acad Sci USA 90: Ivanov S, Liao SY, Ivanova A, Danilkovitch-Miagkova A, Tarasova 1127–1131. N, Weirich G et al. (2001). Expression of hypoxia-inducible cell- Niederer SA, Swietach P, Wilson DA, Smith NP, Vaughan-Jones RD. surface transmembrane carbonic anhydrases in human cancer. Am (2008). Measuring and modeling chloride-hydroxyl exchange in the J Pathol 158: 905–919. Guinea-pig ventricular myocyte. Biophys J 94: 2385–2403. Kallinowski F, Schlenger KH, Runkel S, Kloes M, Stohrer M, Pan P, Leppilampi M, Pastorekova S, Pastorek J, Waheed A, Sly WS Okunieff P et al. (1989). Blood flow, metabolism, cellular micro- et al. (2006). Carbonic anhydrase in CA II-deficient environment, and growth rate of human tumor xenografts. Cancer (Car2-/-) and CA IX-deficient (Car9-/-) mice. J Physiol 571: 319–327. Res 49: 3759–3764. Pastorek J, Pastorekova S, Callebaut I, Mornon JP, Zelnik V, Koukourakis MI, Giatromanolaki A, Sivridis E, Simopoulos K, Opavsky R et al. (1994). Cloning and characterization of MN, a Pastorek J, Wykoff CC et al. (2001). Hypoxia-regulated carbonic human tumor-associated protein with a domain homologous to anhydrase-9 (CA9) relates to poor vascularization and resistance of carbonic anhydrase and a putative helix-loop-helix DNA binding squamous cell head and neck cancer to chemoradiotherapy. Clin segment. Oncogene 9: 2877–2888. Cancer Res 7: 3399–3403. Pastorekova S, Parkkila S, Parkkila AK, Opavsky R, Zelnik V, Kunz-Schughart LA, Doetsch J, Mueller-Klieser W, Groebe K. (2000). Saarnio J et al. (1997). Carbonic anhydrase IX, MN/CA IX: Proliferative activity and tumorigenic conversion: impact on cellular analysis of stomach complementary DNA sequence and expression metabolism in 3-D culture. Am J Physiol Cell Physiol 278: C765–C780. in human and rat alimentary tracts. Gastroenterology 112: 398–408. L’Allemain G, Paris S, Pouyssegur J. (1985). Role of a Na+- Pastorekova S, Pastorek J. (2004) Supuran C, Scozzafava A and dependent Cl-/HCO3- exchange in regulation of intracellular pH in Conway J (eds). Carbonic Anhydrase: its Inhibitors and Activators. fibroblasts. J Biol Chem 260: 4877–4883. CRC Press: Boca Raton, FL. Lagarde AE, Franchi AJ, Paris S, Pouyssegur JM. (1988). Effect of Potter CP, Harris AL. (2003). Diagnostic, prognostic and therapeutic mutations affecting Na+:H+ antiport activity on tumorigenic implications of carbonic anhydrases in cancer. Br J Cancer 89: 2–7. potential of hamster lung fibroblasts. J Cell Biochem 36: 249–260. Pouyssegur J, Dayan F, Mazure NM. (2006). Hypoxia signalling in Lardner A. (2001). The effects of extracellular pH on immune cancer and approaches to enforce tumour regression. Nature 441: function. J Leukoc Biol 69: 522–530. 437–443. Lee AH, Tannock IF. (1998). Heterogeneity of intracellular pH and of Pouyssegur J, Franchi A, L’Allemain G, Paris S. (1985). Cytoplasmic mechanisms that regulate intracellular pH in populations of pH, a key determinant of growth factor-induced DNA synthesis in cultured cells. Cancer Res 58: 1901–1908. quiescent fibroblasts. FEBS Lett 190: 115–119.

Oncogene Carbonic anhydrase IX in tumour pH regulation P Swietach et al 6521 Pouyssegur J, Sardet C, Franchi A, L’Allemain G, Paris S. (1984). A and intracellular in three-dimensional tumor cell growths. J Biol specific mutation abolishing Na+/H+ antiport activity in hamster Chem 284: 20299–20310. fibroblasts precludes growth at neutral and acidic pH. Proc Natl Swietach P, Spitzer KW, Vaughan-Jones RD. (2007). pH-Dependence Acad Sci USA 81: 4833–4837. of extrinsic and intrinsic H(+)-ion mobility in the rat ventricular Romero MF, Fulton CM, Boron WF. (2004). The SLC4 family of myocyte, investigated using flash photolysis of a caged-H(+) HCO 3 – transporters. Pflugers Arch 447: 495–509. compound. Biophys J 92: 641–653. Saarnio J, Parkkila S, Parkkila AK, Haukipuro K, Pastorekova S, Swietach P, Wigfield S, Cobden P, Supuran CT, Harris AL, Vaughan- Pastorek J et al. (1998). Immunohistochemical study of colorectal Jones RD. (2008). Tumor-associated carbonic anhydrase 9 spatially tumors for expression of a novel transmembrane carbonic coordinates intracellular pH in three-dimensional multicellular anhydrase, MN/CA IX, with potential value as a marker of cell growths. J Biol Chem 283: 20473–20483. proliferation. Am J Pathol 153: 279–285. Swietach P, Zaniboni M, Stewart AK, Rossini A, Spitzer KW, Sardet C, Franchi A, Pouyssegur J. (1989). Molecular cloning, primary Vaughan-Jones RD. (2003). Modelling intracellular H þ ion diffu- structure, and expression of the human growth factor-activatable sion. Prog Biophys Mol Biol 83: 69–100. Na+/H+ antiporter. Cell 56: 271–280. Thomas RC. (1989). Cell growth factors. Bicarbonate and pHi Sauvant C, Nowak M, Wirth C, Schneider B, Riemann A, Gekle M response. Nature 337: 601. et al. (2008). Acidosis induces multi-drug resistance in rat Thomlinson RH, Gray KH. (1955). The histological structure of some prostate cancer cells (AT1) in vitro and in vivo by increasing the human lung cancers and the possible implications for radiotherapy. activity of the p-glycoprotein via activation of p38. Int J Cancer 123: Br J Cancer 9: 539–549. 2532–2542. Turner JR, Odze RD, Crum CP, Resnick MB. (1997). MN antigen Sennoune SR, Bakunts K, Martinez GM, Chua-Tuan JL, Kebir Y, expression in normal, preneoplastic, and neoplastic esophagus: a Attaya MN et al. (2004). Vacuolar H+-ATPase in human clinicopathological study of a new cancer-associated biomarker. breast cancer cells with distinct metastatic potential: distribution Hum Pathol 28: 740–744. and functional activity. Am J Physiol Cell Physiol 286: Turner KJ, Crew JP, Wykoff CC, Watson PH, Poulsom R, Pastorek J C1443–C1452. et al. (2002). The hypoxia-inducible genes VEGF and CA9 are Spitzer KW, Skolnick RL, Peercy BE, Keener JP, Vaughan-Jones RD. differentially regulated in superficial vs invasive bladder cancer. Br J (2002). Facilitation of intracellular H(+) ion mobility by CO(2)/ Cancer 86: 1276–1282. HCO(3)(-) in rabbit ventricular myocytes is regulated by carbonic Vasseur M, Cauzac M, Alvarado F. (1989). Electroneutral, HCO3(-)- anhydrase. J Physiol 541: 159–167. independent, pH gradient-dependent uphill transport of Cl- by ileal Sterling D, Reithmeier RA, Casey JR. (2001). A transport metabolon. brush-border membrane vesicles. Possible role in the pathogenesis Functional interaction of carbonic anhydrase II and chloride/ of chloridorrhea. Biochem J 263: 775–784. bicarbonate exchangers. J Biol Chem 276: 47886–47894. Vaughan-Jones RD, Peercy BE, Keener JP, Spitzer KW. (2002). Stewart AK, Kurschat CE, Vaughan-Jones RD, Alper SL. (2009). Intrinsic H(+) ion mobility in the rabbit ventricular myocyte. Putative re-entrant loop 1 of AE2 transmembrane domain has a J Physiol 541: 139–158. major role in acute regulation of anion exchange by pH. J Biol Vaughan-Jones RD, Wu ML. (1990). Extracellular H+ inactivation of Chem 284: 6126–6139. Na(+)-H+ exchange in the sheep cardiac Purkinje fibre. J Physiol Stock C, Mueller M, Kraehling H, Mally S, Noel J, Eder C et al. 428: 441–466. (2007). pH nanoenvironment at the surface of single melanoma Vaupel P, Kallinowski F, Okunieff P. (1989). Blood flow, oxygen and cells. Cell Physiol Biochem 20: 679–686. nutrient supply, and metabolic microenvironment of human tumors: Sun B, Leem CH, Vaughan-Jones RD. (1996). Novel chloride- a review. Cancer Res 49: 6449–6465. dependent acid loader in the guinea-pig ventricular myocyte: part Wohlhueter RM, Plagemann PG. (1981). Hexose transport and of a dual acid-loading mechanism. J Physiol 495(Pt 1): 65–82. phosphorylation by Novikoff rat hepatoma cells as function of Supuran CT. (2008). Carbonic anhydrases: novel therapeutic extracellular pH. J Biol Chem 256: 869–875. applications for inhibitors and activators. Nat Rev Drug Discov Wykoff CC, Beasley NJ, Watson PH, Turner KJ, Pastorek J, Sibtain A 7: 168–181. et al. (2000). Hypoxia-inducible expression of tumor-associated Sutherland RM. (1988). Cell and environment interactions in carbonic anhydrases. Cancer Res 60: 7075–7083. tumor microregions: the multicell spheroid model. Science 240: Xu C, Lo A, Yammanuru A, Tallarico AS, Brady K, Murakami A 177–184. et al. (2010). Unique biological properties of catalytic domain Svastova E, Zilka N, Zat’ovicova M, Gibadulinova A, Ciampor F, directed human anti-CAIX antibodies discovered through phage- Pastorek J et al. (2003). Carbonic anhydrase IX reduces E-cadherin- display technology. PLoS One 5: e9625. mediated adhesion of MDCK cells via interaction with beta-catenin. Zavada J, Zavadova Z, Pastorek J, Biesova Z, Jezek J, Velek J. (2000). Exp Cell Res 290: 332–345. Human tumour-associated cell adhesion protein MN/CA IX: Swietach P, Patiar S, Supuran CT, Harris AL, Vaughan-Jones RD. identification of M75 epitope and of the region mediating cell (2009). The role of carbonic anhydrase 9 in regulating extracellular adhesion. Br J Cancer 82: 1808–1813.

Oncogene