Carbonic Anhydrase II Mediates Malignant Behavior of ... - ATS Journals

ORIGINAL RESEARCH Carbonic Anhydrase II Mediates Malignant Behavior of Pulmonary Neuroendocrine Tumors Yuanxiang Zhou1, Reza Bayat Mokhtari1, Jie Pan1, Ernest Cutz1, and Herman Yeger1,2 1

Division of Pathology, Department of Paediatric Laboratory Medicine, and 2Program in Developmental & Stem Cell Biology, The Research Institute, The Hospital for Sick Children, Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada

Abstract In normal lung, the predominant cytoplasmic carbonic anhydrase (CA) isozyme (CAII) is highly expressed in amine- and peptideproducing pulmonary neuroendocrine cells where its role involves CO2 sensing. Here, we report robust cytoplasmic expression of CAII by immunohistochemistry in the tumor cells of different native neuroendocrine tumor (NET) types, including typical and atypical carcinoids and small-cell lung carcinomas, and in NET and non-NET tumor cell lines. Because, in both pulmonary neuroendocrine cell and related NETs, the hypercapnia-induced secretion of bioactive serotonin (5-hydroxytryptamine) is mediated by CAII, we investigated the role of CAII in the biological behavior of carcinoid cell line H727 and the type II cell–derived A549 using both in vitro clonogenicity and in vivo xenograft model. We show that short hairpin RNA–mediated down-regulation of CAII resulted in significant reduction in clonogenicity of H727 and A549 cells in vitro, and marked suppression of tumor growth in vivo. CAII-short hairpin RNA cell–derived xenografts showed significantly reduced mitosis (phosphohistone H3 marker) and proliferation associated antigen Ki-67 (Ki67 marker), and significantly increased apoptosis by terminal deoxynucleotidyl transferase dUTP nick end labeling assay.

Tumors arising from pulmonary neuroendocrine cells (PNECs), include carcinoids and small-cell lung carcinomas (SCLCs) (1, 2). Pulmonary neuroendocrine tumors (NETs), particularly SCLCs, are highly malignant and resistant to current treatments (3, 4). Similarly, lung carcinoids are difficult to treat, and their prognosis is

Using an apoptosis gene array, we found no association with caspases 3 and 8, but with a novel association of CAII-mediated apoptosis with specific mitochondrial apoptosis–associated proteins. Furthermore, these xenografts showed a significantly reduced vascularization (CD31 marker). Thus, CAII may play a critical role in NET lung tumor growth, angiogenesis, and survival, possibly via 5-hydroxytryptamine, known to drive autocrine tumor growth. As such, CAII is a potential therapeutic target for the difficult-to-treat lung NETs. Keywords: carbonic anhydrase II; tumor; clonogenicity; apoptosis; vascularization

Clinical Relevance The cytoplasmic carbonic anhydrase (CA) II is abundantly expressed in many types of cancers. Suppression of CAII resulted in significantly reduced clonogenicity and tumorigenicity. Thus, CAII could be a novel target for lung cancer treatment. In addition, targeting CAII could reveal additional tumor associated apoptosis pathways.

variable (5). Therefore, a greater knowledge on NET cell biology is imperative to improve the outcome of patients with these tumors. NETs are derived from solitary PNECs and/or innervated PNEC clusters, called neuroepithelial bodies (NEBs) that, in normal lung, function as airway chemosensors for O2 and CO2 (6, 7). The O2

chemosensory mechanism in NEB cells has been well characterized, and consists of a reduced nicotinamide adenine dinucleotide phosphate oxidase complex linked to O2-sensitive potassium voltage-gated (Kv1) channels (8, 9). When triggered by hypoxia, PNECs/NEB cells release potent bioactive substances, including 5-hydroxytryptamine

( Received in original form February 18, 2014; accepted in final form June 25, 2014 ) This work was supported by Canadian Institutes for Health Research grant MOP-15270 and a Cancer Research Society/Canadian Neuroendocrine Tumor Society joint grant (H.Y. and E.C.). Correspondence and requests for reprints should be addressed to Herman Yeger, Ph.D., Division of Pathology, Department of Paediatric Laboratory Medicine, The Hospital for Sick Children, 555 University Avenue, Toronto, ON, M5G1X8 Canada. E-mail: [email protected] This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org Am J Respir Cell Mol Biol Vol 52, Iss 2, pp 183–192, Feb 2015 Copyright © 2015 by the American Thoracic Society Originally Published in Press as DOI: 10.1165/rcmb.2014-0054OC on July 14, 2014 Internet address: www.atsjournals.org

Zhou, Mokhtari, Pan, et al.: CAII Mediates Malignant Behavior of NETs

183

GAPDH

CAXIV

CAXIII

CAXII variant 1

CAXII variant 2

CAX

CAXI

CAIX variant 1

CAIX variant 2

CAVIII

CAVII variant 2

CAVI

CAVb

CAIV

CAVa

CAIII

H727

CAII

CAI

A

CAVII variant 1

ORIGINAL RESEARCH

NS-shRNA CAII-shRNA A549 NS-shRNA

β-actin CAII

H727

D

E

C CAII percentages over NS-shRNA

CAII-shRNA

NS-shRNA

B

CAII-shRNA

NS-shRNA

CAII-shRNA

A549

120 100 80 60 40 ***

20 0

***

NS-shRNA CAll-shRNA NS-shRNA CAll-shRNA H727 A549

H727

CAII

DAPI

CAII

DAPI

A549

Figure 1. Carbonic anhydrase (CA) gene expression and CAII protein in H727 and A549 with and without CAII–short hairpin RNA (shRNA) treatment. (A) RT-PCR showing the CA expression profile in nonspecific (NS)- and CAII-shRNA H727 and A549 cells. Primer sequences and amplicon sizes (bp) are listed in Table 1. (B) Western blot analysis showing CAII suppression by CAII-shRNA in both H727 and A549 cells. Protein loading was normalized by b-actin blotting. (C) Percentages of CAII protein levels in CAII-shRNA cells over NS-shRNA cells as normalized to b-actin intensities. ***P , 0.001 versus the NS-shRNA control. (D and E) Immunofluorescence showing cytoplasmic localization of CAII (red) in H727 (D) and A549 (E) cells. Nuclei (blue) were stained with 49,6-diamidino-2-phenylindole (DAPI). Scale bars, 25 mm. GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

184

American Journal of Respiratory Cell and Molecular Biology Volume 52 Number 2 | February 2015

ORIGINAL RESEARCH Table 1. RT-PCR Primers for the Amplification of Carbonic Anhydrase Messenger RNAs

Forward Primer

Reverse Primer

catggttcagaacatacagtgga cataaggacttccccattgccaa ctggcatgaacttttcccaaatgcc cgggagcccattcagcttcac agagcacacagtggacggccac ccaccagccaagcaacccccta cttctgctgtccctgttcctg aagagagctccttcatcctgcc ggacggcagcggcatgaccg aatcagttctttcgggaggacc ggctccctgaagttagaggatct agactcaggtcccaggactggac cgttttcccgtcgagtgatgcac acagaggagctgagagagggg cgctgaatattaccgctaccgg cgctgaatattaccgctaccgg gccattccttcaatgttgactttg ggcagtacttccgctacaatggc catggcaaattccatggcaccg

gggtcaaaatttgtgaatggggc atcaagtgaaccccagtgaaag gtgctcagagccatgatcatccg tgcagaagtgagccatcatcgc gcagggagtcaggaggggtcgaag gcagcagagcaagacctagtctc cctgggtctcatagcctgtcat agtcaccacggttcggtcatcg agtcaccacggttcggtcatcg gccttcaagccaacatgttcctt gcagggtcagttgcacactgtgg ccaaaaaccagggctaggatgtc ctatgttgactggcgactgccg gcttccccacagcacacagac gtgagaggatgatgccttgggag cctgccgcagtacagacttgca tgggaattaggttcaccaatctg cagccaaccaagattcctacacc cattgctgatgatcttgaggctg

Amplicon (bp)

treatment, cells were maintained in the incubator containing a gas mixture consisting of 1% O2 and 5% CO2 balanced with N2. shRNA-Mediated Knockdown of CAII

CAI CAII CAIII CAIV CAVa CAVb CAVI CAVII variant 1 CAVII variant 2 CAVIII CAIX variant 1 CAIX variant 2 CAX CAXI CAXII variant 1 CAXII variant 2 CAXIII CAXIV GAPDH

235 255 277 239 242 278 208 270 259 256 304 349 256 261 245 252 280 283 282

Definition of abbreviations: CA, carbonic anhydrase; GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

The shRNA clones that expressed nonspecific (NS) or CAII-shRNAs were obtained from Open Biosystems (Lafayette, CO). Three CAII-shRNAs (clones V3LHS351110, V3LHS-351107, and V3LHS351109; the latter two were for the purpose of excluding the off-target possibility) that suppressed CAII at a similar level were used. The plasmids prepared from the clones were transfected into the H727 and A549 cells using the Lipofectamine 2,000 Transfection Reagent (Invitrogen, Burlington, ON, Canada). The stably transfected cells after puromycin (1.5 mg/ml) selection were denoted either NS-shRNA or CAII-shRNA cells, depending on the type of shRNA transfected. Clonogenic Assay

(5-HT; serotonin), and peptides, such as bombesin (6). It is thought that CO2 sensing in chemoreceptor cells is predominantly performed by carbonic anhydrases (CAs) (10, 11). We have found recently that, in addition to hypoxia, hypercapnia and acid pH also trigger 5-HT secretion from NEBs, suggesting that NEBs are polymodal airway sensors involved in both O2 and CO2 chemosensing in pulmonary airways (12). Previously, we profiled and compared the expression of CA genes in native PNECs/NEB cells isolated from human neonatal lung versus adjacent normal airway epithelium and the PNEC-derived lung carcinoid tumor cell line, H727, as well as tumor cell line, A549, derived from lung adenocarcinoma. We found several differences between native PNECs/NEBs and the airway epithelium, with PNECs/ NEBs expressing several CA genes, including CAII. The mRNA profiles in PNECs/NEBs closely resembled those in H727 (12) and A549 (unpublished data), with all of them showing strong CAII expression. In this study, we investigated the expression of CAII in different lung cancers and studied the role of CAII in the biology of NETs derived from PNECs (cell line H727) and the alveolar type 2 cells (cell line A549) representative of

a non-NET tumor. Here, we report that suppression of CAII activity by CAII–short hairpin RNAs (shRNAs) in H727 and A549 leads to a markedly reduced clonogenicity in vitro and tumorigenicity in vivo. The results portend a potential therapeutic approach for pulmonary NETs as well as other lung tumors.

Materials and Methods Cell Culture

The typical carcinoid cell line, NCI-H727 (ATCC CRL-5815; American Type Culture Collection, Manassas, VA), the atypical carcinoid cell line, NCI-H720 (ATCC CRL-5838), a third bronchial carcinoid cell line, NCI-H835 (ATCC CRL-5843), and the type II cell–derived carcinoma cell line, A549 (ATCC CCL185), were cultured in RPMI-1640 supplemented with 10% FBS and antibiotics at 378 C and 5% CO2, as previously described (13). The lung adenocarcinoma cell line, NCI-H23, large-cell lung carcinoma cell line, NCIH460, and squamous cell lung carcinoma cell line, NCI-H520, were kindly provided by E. Diamandis (Mt. Sinai Hospital, Toronto, ON, Canada), and cultured under similar conditions. For hypoxia


Equal volumes of MethoCult matrix (Stem Cell Technologies, Vancouver, BC, Canada) and 23 RPMI-1640 medium, supplemented with 23 FBS and 23 antibiotics, were mixed. H727 and A549 monolayers were dissociated with 0.25% trypsin, washed twice with 13 PBS (pH 7.4), and resuspended in RPMI-1640 medium at a density of 5 3 104 cells/ml. The resuspended cells (100 ml) were thoroughly mixed with MethoCult/RPMI mixture, seeded in a 35-mm suspension tissue culture dish, and incubated at 378 C and 5% CO2. Twice per week, an equal volume of fresh medium was added on top of the solidified matrix, and was left to equilibrate for 2 hours before being carefully removed. At different time points, the numbers and the sizes of colonies were assessed under the microscope. Xenograft Assay

A total of 1.5 million NS-shRNA– and CAII-shRNA–treated H727 and A549 cells were subcutaneously injected into the inguinal fat pads of nude mice. Tumors were measured during the course of tumor initiation and growth, and tumor volumes were calculated using the formula: volume = (width)2 3 length/2. Finally, the mice were killed and tumors extirpated for analysis. 185

ORIGINAL RESEARCH Terminal Deoxynucleotidyl Transferase dUTP Nick End Labeling Assay

The terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay was performed on 5-mm sections prepared from formalin-fixed, paraffin-embedded xenografts, using the In Situ Cell Death Detection Kit (Roche, Laval, PQ, Canada) and the protocol recommended by the manufacturer, except that the positive control was treated with 500 U/ml DNaseI (Roche) before adding the TUNEL reaction buffer. The peroxidase reaction was performed with stable diaminobenzidine solution (Invitrogen). Finally, the slides were counterstained with hematoxylin and examined under light microscopy. See the MATERIALS AND METHODS section in the online supplement for descriptions of RT-PCR, real-time PCR analysis, Western blot analysis, immunofluorescence, immunohistochemistry, mitotic index, and statistical analysis.

was seen in H727 and H720, and approximately equal amounts in the other cell lines, but at a lower level of expression. CA profiling of CAs I–XIV and variants was conducted by RT-PCR on H727 and A549 cells (Figure 1A and Table 1). Both cell lines expressed a selected number of CAs. CAII mRNA was found to be abundantly expressed in both H727 and A549 cells (Figure 1A). At the protein level, CAII expression was found in both cell lines, but was significantly reduced in the CAIIshRNA–treated cells. In H727 cells, CAII expression was almost completely abolished by CAII-shRNAs. In A549

A

cells, the CAII protein level was also significantly reduced by CAII-shRNAs (P , 0.001; Figures 1B and 1C). Note also the significant reduction of CAII mRNA after CAII-shRNA treatment (Figure 1A). Thus, knockdown of CAII significantly affected CAII expression in both tumor cell types. By immunofluorescence labeling, the CAII protein was localized diffusely in the cytoplasm of both tumor cell lines (Figures 1D and 1E). CAII-shRNA–Mediated Knockdown of CAII Affects Tumor Cell Behavior

CAII-shRNA–mediated CAII knockdown profoundly affected H727 cells that were

NS-shRNA

CAII-shRNA

1 hr attachment

ovrnight attachment

Results Expression of CAII in Pulmonary NETs and Other Lung Cancer Tumors and Cell Lines: CAII-shRNA DownRegulates CAII Expression

The presence of CAII was demonstrated by immunohistochemistry in native, surgically resected human NETs, including typical and atypical carcinoids and SCLCs (see Figure E1 in the online supplement). Figure E1 shows the hematoxylin and eosin histopathology and the CAII immunostaining, as correlated with immunostaining for the neuroendocrine secretory granule marker, chromogranin A, that confirms the NET phenotype. An abundant expression of CAII was found in the cytoplasm of tumor cells, but not in the surrounding stromal component. Human kidney served as the positive control, and no staining was obtained after omission of the primary anti-CAII antibody (data not shown). Expression of CAII was confirmed in the panel of bronchial carcinoids cell lines (H720, H727, and H835), as well as in the three non-NET lung cancer cell lines (H23, H460, and H520) by Western blot analysis (Figure E2). Strong expression 186

B

48 hr NOX

48 hr NOX + 30 min in PBS

C

48 hr HOX

48 hr HOX + 30 min in PBS

Figure 2. CAII-shRNA transfected H727 cells were slower in attachment and faster in detachment, indicating altered adhesion properties. (A) Attachment of NS- and CAII-shRNA cells. Monolayer cells were trypsinized, washed in PBS, resuspended in RPMI-1640 medium, and seeded in culture plates. Attached cells were photographed after 1 hour and overnight. Note that after 1 hour the NS-shRNA cells were attached and well spread, whereas the CAII-shRNA cells were still rounded. (B) Detachment of NS- and CAII-shRNA cells after 48 hours in normoxia (NOX). Newly seeded cells were cultured in normoxia condition for 48 hours, and then the culture medium was replaced with 13 PBS. No obvious difference in detachment was observed at 30 minutes after medium replacement. (C) Detachment of NS- and CAII-shRNA cells after 48 hours in hypoxia (HOX) culture condition. Newly seeded cells were cultured in HOX (1% O2) condition for 48 hours, and then the medium was replaced by 13 PBS. After 30 minutes in 13 PBS, the NS-shRNA cells were relatively well spread, whereas the CAII-shRNA–treated cells became obviously rounded and appeared ready to detach.


ORIGINAL RESEARCH

Given the strong effect of CAII suppression on clonogenicity in vitro, we surmised that CAII suppression could have a similar inhibitory effect in vivo. To test this possibility, 1.5 3 106 shRNA cells were injected subcutaneously into immunocompromised nude mice. The tumor sizes were measured regularly on the days after inoculation. The CAIIshRNA cells produced dramatically smaller tumor masses than the NSshRNA cells for both cell lines, and such

CAII-shRNA–expressing plasmids were transfected into H727. When later xenografted into the nude mice, similar results as shown in Figures 4A to 4E were observed (Figure E3). Western blot analysis revealed the continued, near-complete depletion of CAII protein in the H727 CAIIshRNA tumors, and significant suppression of CAII protein in the A549 CAII-shRNA tumors (Figures 4F and 4G; P , 0.001), similar to that in the CAII-shRNA cells (Figures 1B and 1C). Thus, CAII suppression in H727 and A549 cells significantly inhibited tumor growth in a mouse xenograft model.

C 4000 3500 3000 2500 2000 1500 1000 500 0

A549

H727

H727

*** NS-shRNA

CAII-shRNA CAII-shRNA

NS-shRNA

B

D

1400

A549

1200 1000 800 600 400 ***

200 0 NS-shRNA

E

140 120

Length

CAII-shRNA

F

H727

Width

100 80

**

60

** **

40

**

20 0 NS-shRNA CAII-shRNA NS-shRNA CAII-shRNA N=18

N=18

N=18

Day 15

N=12 Day 30

Colony size in µm

CAII-shRNA–Mediated Knockdown of CAII Causes a Marked Reduction in the Growth of Tumor Xenografts In Vivo

A # of colonies/plate

Suppression of the membrane-bound CAIX, a tumor marker associated with tumor metastasis, has been shown to reduce tumor growth both in vitro and in vivo, an effect that was thought to be contributed partly by the CAIX protein in counteracting acidosis through the regulation of intracellular pH (pHi) (14). Because cytosolic CAII is directly involved in pHi regulation, it may also be involved in the regulation of tumor cell growth. To test this possibility, the shRNA-treated H727 and A549 cells were first assessed using a clonogenic assay. The CAII-shRNA transfected cells produced not only significantly fewer colonies (Figures 3A–3C; P , 0.001 for both cell lines), but also much smaller colonies (Figures 3E and 3F; P , 0.005 on Days 15 and 30 for H727; P , 0.01 on Day 20 and P , 0.005 on Day 30 for A549). Figure 3D shows the phenotype of a typical colony. The fold difference in colony numbers was greater between the NS-shRNA and CAII-shRNA transfected H727 cells than between the two similarly treated A549 cells (Figures 3A and 3B). These data demonstrate the strong inhibitory effect of CAII suppression on clonogenic capacity.

# of colonies/plate

CAII-shRNA–Mediated Knockdown of CAII Affects Tumor Cell Clonogenicity In Vitro

differences became more pronounced with each day after inoculation (Figures 4A and 4B; P , 0.001). Whereas NS-shRNA transfected cells showed exponential tumor growth, the CAIIshRNA cells exhibited a highly reduced growth rate within the first few days after inoculation. Extirpated CAII-shRNA xenografts were paler and appeared less vascularized than the large NS-shRNA tumors (Figures 4C and 4D), and weighed significantly less than their NS-shRNA counterparts (Figure 4E; P , 0.001). To exclude the possibility of shRNA-related off-target effects, two more

Colony size in µm

slow to attach and spread in between passages, and faster to detach after hypoxia treatment (1% O2), suggesting alterations in membrane proteins involved in cell adhesion and spreading (Figures 2A–2C). These effects were not observed in CAII-suppressed A549 cells (data not shown).

200 180 160 140 120 100 80 60 40 20 0

A549

Length Width

** **

* *

NS-shRNA CAII-shRNA NS-shRNA CAII-shRNA N=19

N=12 Day 20

N=18

N=13 Day 30

Figure 3. CAII-shRNA cells had reduced clonogenic capacity in vitro. (A) Number of colonies formed from H727 cells on Day 15, representative of three duplicates. (B) Number of colonies formed from A549 cells on Day 15, representative of three duplicates. (C) Photos of 4-mm2 area showing the difference in the number of colonies formed from the NS-shRNA cells (left panel) and the CAII-shRNA cells (right panel). Scale bars, 500 mm. (D) Phase–contrast micrograph of a colony 30 days after cell seeding. Scale bar, 50 mm. (E) Sizes of colonies formed from H727 cells on Days 15 and 30. (F) Sizes of colonies formed from A549 cells on Days 20 and 30. N, number of colonies measured. **P , 0.005 and *P , 0.01 versus the NS-shRNA control.


187

ORIGINAL RESEARCH B

H727

8

NS-shRNA

7

CAII-shRNA

6 5 4

4

3 2

A549

NS-shRNA CAII-shRNA

3

***

***

3 Tumor volume (cm )

9

3 Tumor volume (cm )

A

2

1

1 0

0 0

3

6

9

12

0

5

10

15

Days

Days

C

D

NS-shRNA

NS-shRNA

CAII-shRNA

CAII-shRNA A549

H727

E

7

Tumor weight (g)

6 5 4 3 2 ***

1

*** 0 NS-shRNA

CAII-shRNA

NS-shRNA

CAII

β-actin

A549

G

120

CAII percentages over NS-shRNA

CAII-shRNA

NS-shRNA

NS-shRNA

F

CAII-shRNA

H727

100 80 60 40

A549

***

20 0

H727

CAII-shRNA

***

NS-shRNA CAII-shRNA NS-shRNA CAII-shRNA H727 A549

Figure 4. The tumorigenicity of CAII-shRNA H727 and A549 cells was significantly reduced in vivo. (A and B) Significant difference in volumes of tumors derived from NS- and CAII-shRNA H727 and A549 cells, after subcutaneous inoculation of 1.5 3 10 6 cells into immunocompromised mice. ***P , 0.001 versus the NS-shRNA control. (C and D) Visual differences in sizes between the NS-shRNA and CAII-shRNA tumors developed from H727 and A549 cells. Scale bars, 1 cm. (E) Significant difference in tumor weights between the NS-shRNA and CAII-shRNA tumors. ***P , 0.001 versus the NS-shRNA control. (F) Western blot analysis showing the difference in CAII protein levels between the NS-shRNA and CAII-shRNA tumors. (G) Percentages of CAII protein in the CAII-shRNA tumors over that in the NS-shRNA tumors. Experimental values were normalized against the b-actin intensity. ***P , 0.001 versus the NS-shRNA control.

CAII-shRNA–Mediated Knockdown of CAII Induces Apoptosis of Tumor Cells and Decreases Tumor Cell Proliferation and Tumor Angiogenesis

To gain further insight into the marked differences in tumor size, we examined the degree of apoptosis using the TUNEL assay. In tumors derived from both H727 and A549 188

cells, the percentages of TUNEL-positive cells in the CAII-shRNA tumors were significantly higher than those in the control (Figures 5A–5C; P , 0.001). Although a small degree of apoptosis is normally expected in tumors, the significant increase in the CAII-shRNA tumors suggested that suppression of CAII compromises tumor cell survival. To

gain further insight as to the apoptotic mechanisms, real-time PCR analyses were performed using an apoptotic gene array (see Table E1). Overt differences in the activities of caspases 2, 3, 7, 8, and 9 were not found between the CAII-and the NS-shRNA tumors. However, in the CAII-shRNA tumors, there was a 10-fold decrease in the expression of the apoptosis inhibitor, HRT (also known as SAG/ ROC/Rbx/Hrt-sensitve to apoptosis/regulator of cullins/ring box protein/Hrt), and a fourfold increase in BH3 interacting-domain death agonist, a proapoptotic protein acting at the mitochondrial membrane. Most strikingly, in the CAII-shRNA tumors there was a 128-fold increase in expression of lymphotoxin-b receptor, which has been proposed to initiate mitochondriamediated apoptosis by an as-yet unknown mechanism (15). In addition, we studied the AKT-mediated survival signaling pathway that can trigger apoptosis, by examining the levels of phospho-Akt (Akt, also known as phospho-protein kinase B) and phospho– phosphoinositide-dependent kinase 1. Both the phosphorylated Akt and phosphoinositide-dependent kinase 1 were significantly reduced in the CAII-shRNA tumors (Figure E4), suggesting that this pathway was affected. Thus, these data suggest a potentially new mechanism for triggering apoptosis via loss of CAII activity. Next, we examined the other side of the equation—namely, cell proliferation. We assessed cell proliferation marker proteins, proliferation-associated antigen Ki-67 and mitosis marker phosphohistone H3 (PH3), using immunohistochemical methods. The tumors derived from the CAII-shRNA cells showed significantly reduced Ki-67 (Figures 6A and 6B; P , 0.001) and PH3 (Figures 6C and 6D; P , 0.001) labeling, indicating a marked reduction in cell proliferation. The mitotic index was obtained based on enumeration of PH3-positive cells. Figure 6D shows that the mitotic index was obviously reduced in the CAII-shRNA tumors derived from both the H727 and A549 cells. Therefore, the CAII-shRNA mediated knockdown in H727 and A549 tumors had significant effects on tumor cell proliferative capacity. As additional supporting data, Western blot analysis was performed to evaluate the PH3 levels. Compared with the controls,


ORIGINAL RESEARCH A % of TUNEL-positive cells

12

***

10 8 6 *** 4 2 0

NS-shRNA

CAll-shRNA

NS-shRNA

H727

CAll-shRNA A549

H727

B NS-shRNA

CAII-shRNA

the PH3 protein was reduced by 45.7% in H727 and by 67.3% in A549 CAIIshRNA tumors (Figures 6E and 6F; P , 0.001). One other factor possibly contributing to the growth inhibition could be the reduced ability of tumors to elicit vascularization, as suggested by the palor of the extirpated tumors on gross appearance. To examine the vascularization of the xenograft tumors, we used immunohistochemistry for CD31, a marker of endothelial cells. We observed that, in control NS-shRNA–treated tumors, CD31 immunopositive cells appeared more abundant in terms of frequency and signal intensity (Figure 7A). At the protein level, the expression of CD31 was reduced about threefold in the H727 CAIIshRNA–treated tumors compared with the control NS-shRNA tumors. The reduction in CD31 protein expression was also observed in A549 CAII-shRNA tumors (Figures 7B and 7C). Thus, the CAIIshRNA–mediated knockdown of tumors leads to significantly reduced angiogenic capacity.

Discussion Negative

Positive

TUNEL

A549

C NS-shRNA

CAII-shRNA

Negative

Positive

TUNEL

Figure 5. The CAII-shRNA tumor xenografts show an increased level of apoptosis. (A) Percentages of apoptotic cells over the total number of cells in both NS-shRNA and CAII-shRNA tumors. Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL)–positive cells in 12 microscopic fields were examined. ***P , 0.001 versus the NS-shRNA control. (B and C) Immunohistochemistry showing TUNEL-positive and -negative cells in xenograft tumors developed from H727 and A549 cells. “Negative” control represents no terminal deoxynucleotidyl transferase in the TUNEL reaction; and “Positive” control represents DNaseI treatment before the TUNEL reaction. Cells were counterstained with hematoxylin. Scale bars, 50 mm.


The tumor microenvironment plays an important role in tumor biology (16). CAII catalyzes the reversible conversion of CO2 1 H2O ↔ HCO32 1 H1 in cytosol, thus directly regulating pHi, an important component of the cellular microenvironment. To maintain the pH homeostasis on both sides of the plasma membrane, the CAII must work coordinately with such pH homeostasis participants as ion exchange and transportation (17–20), regulators of CA activities (21), and membrane-bound CAs. Often in tumors, a shift in the pH homeostasis will result in an altered microenvironment that either promotes or inhibits carcinogenesis (22). Of all the mammalian CA isozymes, the membrane-bound CAIX has drawn the greatest attention, as it is frequently associated with the malignant and metastatic phenotype (14). The CAIX helps to establish an alkaline pHi favorable for tumor growth and survival, and an acidic extracellular pH favoring tumor cell invasion and metastasis (21). CAIX overexpression has resulted in increased malignant behavior of cancer cells (23), and 189

ORIGINAL RESEARCH A

B

Ki67 immunolabelling

Percentages of Ki-67-positive cells

H727

A549

NS-shRNA

C

50 45 40 35 30 25 20 15 10 5 0

***

NS-shRNA CAII-shRNA H727

CAII-shRNA

D

PH3 immunolabelling

***

H727

NS-shRNA CAII-shRNA A549

50

# of PH3-positive cells

45

A549

40 35 30

***

25 20 15 ***

10 5 0

NS-shRNA

NS-shRNA

CAII-shRNA

CAII-shRNA

PH3 β-actin

NS-shRNA

CAII-shRNA

A549

F 120 % over NS-shRNA PH3

CAII-shRNA

NS-shRNA

NS-shRNA

E

CAII-shRNA

H727

100 80 ***

60

***

40 20 0 NS-shRNA

H727

A549

CAII-shRNA

H727

NS-shRNA

CAII-shRNA

A549

Figure 6. The CAII-shRNA xenograft tumors show a reduced cell proliferation. (A) Proliferationassociated antigen Ki-67 (Ki67) immunohistochemistry. Scale bars, 100 mm. (B) The CAII-shRNA tumors from both H727 and A549 had reduced percentages of Ki67 positive cells. ***P , 0.001 versus the NS-shRNA control. (C) PH3 immunohistochemistry. Scale bars, 100 mm. (D) The CAII-shRNA tumors from both H727 and A549 had reduced number of phosphohistone H3 (PH3)-positive cells. ***P , 0.001 versus the NS-shRNA control. (E) Western blot analysis of mitotic marker PH3 in the xenograft tumors with b-actin as loading control. (F) The percentages of PH3 protein signals in CAII-shRNA tumors relative to those in the NS-shRNA tumors as normalized to b-actin intensities. ***P , 0.001 versus the NS-shRNA control.

CAIX inhibition has brought about the opposite effects (24). A major effort is underway to develop CAIX-targeted therapeutic approaches (25, 26). In contrast to CAIX, the role of the cytoplasmic localized CAII in cancer biology, based on the analysis of clinical samples, has been inconclusive. On the one hand, the CAII levels were found to be negatively correlated to the initiation and severity of colorectal cancer (27) and gastrointestinal stromal tumor (28). On the other hand, in many other tumor types, 190

CAII levels positively correlated with more advanced clinical stages of oral squamous cell carcinoma (29), renal cell carcinoma (30), chromophobe renal cell carcinoma (31), glial tumors (32), and uterine tumor (33). Inhibitor targeting CAII resulted in reduced invasion of renal cancer cells (30). Moreover, studies with 1.3.4-thiadiazole-2sulfonamide that possesses extremely high CAII affinity (34) and the antiangiogenesis role of CAII autoimmune antibody in patients with cancer (35) have indicated that targeting CAII may be associated with

good clinical outcome. We have also recently shown that acetazolamide affected the growth, survival, and tumorigenicity of lung carcinoid cell line, H727 (36). To confirm and validate our findings, the tumor cell lines were individually transfected with three CAII-shRNA plasmids that target different locations in the CAII mRNA. All three CAIIshRNA–expressing cells produced significantly suppressed tumorigenicity in vivo, thus ruling out the possibility of an off-target effect. It is of interest to note that the small tumors derived from CAII-deficient H727 cells showed a notably higher level of CAIX mRNA and CAIX protein expression, as well as a higher level of HIF1a protein compared with controls (data not shown). This suggests that hypoxia resulting from reduced vascularization, as shown in Figure 7, up-regulated CAIX expression. In the context of tumors generated from H727, the increase in CAIX expression alone was insufficient to support tumor growth, as also observed by others (24). Whether CAII and CAIX act cooperatively or independently requires further investigation. It is well established that both native PNEC/NEB cells and H727 cells release 5HT in response to hypoxia and hypercapnia as well as acid pH (13, 37). We have previously reported that CAII suppression significantly reduced 5-HT release from H727 cells under both hypoxia and hypercapnia, as well as acidic (pH 6.8) and alkaline (pH 8.0) conditions (12). This is significant, because 5-HT is a potent growth factor for some human cancers, including human hepatocellular carcinoma (38) and carcinoids that express 5-HT receptors (39). In addition, Nocito and colleagues (40) have shown that colon cancer cells injected into 5-HT–deficient (Tph12/2) mice resulted in reduced tumor growth and microvascularization. This reduction was reversed to near wild-type levels by 5-HT reloading. Whether or not the reduced growth of H727 CAII-shRNA tumors was due in part to the CAII-mediated reduction in 5-HT release remains to be investigated. The CAII-shRNA tumors showed significantly increased apoptosis. However, apoptosis array analysis did not reveal significant differences in the expression levels of key apoptosis-relevant caspases.


ORIGINAL RESEARCH CD31 labelling

A H727

A549

CD31 β-actin

CAII-shRNA

C % over NS-shRNA CD31

CAII-shRNA

NS-shRNA

NS-shRNA

B

CAII-shRNA

NS-shRNA 120 100

***

80 60 40

***

20 0 NS-shRNA CAII-shRNA NS-shRNA CAII-shRNA

H727

A549

H727

A549

Figure 7. CAII-shRNA xenograft tumors are significantly less well vascularized. (A) Vascularization of the tumors was shown by CD31 immunohistochemistry. Scale bars, 200 mm. Arrows point to typical immunolabeling signals representing the vessels. Note that not all the immunolabeled structures are pointed out by arrows. (B) Western blot analysis of vascularization marker, CD31, in the xenograft tumors with b-actin as loading control. (C) The percentages of CD31 signals in CAII-shRNA tumors relative to those in the NS-shRNA tumors as normalized to b-actin intensities. ***P , 0.001 versus the NS-shRNA control.

In addition, Western blot analysis with different antibodies failed to detect caspase 3 in both the CAII- and the NS-shRNA tumors. Thus, the increased apoptosis could be caspase independent. The increased

expression of proapoptotic lymphotoxin-b receptor and BH3 interacting-domain death agonist, and decreased expression of antiapoptotic HRK, in the CAII shRNA tumors may indicate the mitochondrial

References 1. Gorshtein A, Gross DJ, Barak D, Strenov Y, Refaeli Y, Shimon I, Grozinsky-Glasberg S. Diffuse idiopathic pulmonary neuroendocrine cell hyperplasia and the associated lung neuroendocrine tumors: clinical experience with a rare entity. Cancer 2012;118:612–619. 2. Song H, Yao E, Lin C, Gacayan R, Chen MH, Chuang PT. Functional characterization of pulmonary neuroendocrine cells in lung development, injury, and tumorigenesis. Proc Natl Acad Sci USA 2012;109:17531–17536. 3. Kraus AC, Ferber I, Bachmann SO, Specht H, Wimmel A, Gross MW, Schlegel J, Suske G, Schuermann M. In vitro chemo- and radioresistance in small cell lung cancer correlates with cell adhesion and constitutive activation of AKT and MAP kinase pathways. Oncogene 2002;21:8683–8695. 4. Triller N, Korosec P, Kern I, Kosnik M, Debeljak A. Multidrug resistance in small cell lung cancer: expression of P-glycoprotein, multidrug resistance protein 1 and lung resistance protein in chemo-naive patients and in relapsed disease. Lung Cancer 2006; 54:235–240.

involvement of the apoptosis seen in this study (15). Data in this study clearly show that CAII is a target of therapy for lung tumors characteristic of H727 and A549. CAII is also present in other nonneuroendocrine human cancer cell lines, and the neuroendocrine human SCLCs and carcinoids. Thus, targeting CAII could be a valuable approach in the treatment of lung NETs, especially carcinoids, given the fact that we examined most of the available lung carcinoid cell lines. Despite extensive studies, the cure for cancer remains elusive. Recent efforts have focused on metabolic aspects that differentiate tumor cells from normal cells, including molecular pathways governing abnormal glycolysis, high lactic acid production, proton accumulation, as well as reversed pH gradients in which proton dynamics are involved, allowing tumor cells to adapt to the microenvironments that are toxic to normal cells (41). Further studies are required to determine the frequency of CAII expression in various types of lung cancer to validate this enzyme as a potential biological marker and therapeutic target in a wide range and/or in specific subtypes of lung tumors, especially the difficult-totreat NETs. n Author disclosures are available with the text of this article at www.atsjournals.org. Acknowledgments: The authors thank Dr. E. Diamandis (Mt. Sinai Hospital, Toronto) for kindly providing the H23, H460, and H520 cell lines, and Dr. J.C. Cutz (St. Joseph Hospital and McMaster University, Hamilton, Ontario) for providing sections of native lung tumors for immunohistochemical analysis.

5. Noel-Savina E, Descourt R. Focus on treatment of lung carcinoid tumor. Onco Targets Ther 2013;6:1533–1537. 6. Cutz E, Yeger H, Pan J. Pulmonary neuroendocrine cell system in pediatric lung disease-recent advances. Pediatr Dev Pathol 2007;10:419–435. 7.DomnikNJ,CutzE.Pulmonaryneuroepithelialbodiesasairwaysensors:putative role in the generation of dyspnea. Curr Opin Pharmacol 2011;11:211–217. 8. Fu XW, Nurse C, Cutz E. Characterization of slowly inactivating KVa current in rabbit pulmonary neuroepithelial bodies: effects of hypoxia and nicotine. Am J Physiol Lung Cell Mol Physiol 2007;293: L892–L902. 9. Cutz E, Pan J, Yeger H. The role of NOX2 and “novel oxidases” in airway chemoreceptor O(2) sensing. Adv Exp Med Biol 2009;648:427–438. 10. Chandrashekar J, Yarmolinsky D, von Buchholtz L, Oka Y, Sly W, Ryba NJ, Zuker CS. The taste of carbonation. Science 2009;326:443–445. 11. Frommer WB. Biochemistry. CO2mmon sense. Science 2010;327:275–276. 12. Cutz E, Zhou YX, Pan J, Yeger H. Pulmonary neuroepithelial bodies (NEB) are multimodal airway sensors-evidence for CO2/H1 sensing [abstract]. The XVIIIth Meeting of the International Society for Arterial Chemoreception, Hamilton, ON, Canada. 2011;38–39.


191

ORIGINAL RESEARCH 13. Pan J, Bear C, Farragher S, Cutz E, Yeger H. Cystic fibrosis transmembrane conductance regulator modulates neurosecretory function in pulmonary neuroendocrine cell–related tumor cell line models. Am J Respir Cell Mol Biol 2002;27:553–560. 14. Chiche J, Ilc K, Laferri ere ` J, Trottier E, Dayan F, Mazure NM, Brahimi-Horn MC, Pouyssegur ´ J. Hypoxia-inducible carbonic anhydrase IX and XII promote tumor cell growth by counteracting acidosis through the regulation of the intracellular pH. Cancer Res 2009;69:358–368. 15. Wolf MJ, Seleznik GM, Zeller N, Heikenwalder M. The unexpected role of lymphotoxin beta receptor signaling in carcinogenesis: from lymphoid tissue formation to liver and prostate cancer development. Oncogene 2010;29:5006–5018. 16. Swartz MA, Iida N, Roberts EW, Sangaletti S, Wong MH, Yull FE, Coussens LM, DeClerck YA. Tumor microenvironment complexity: emerging roles in cancer therapy. Cancer Res 2012;72:2473–2480. 17. Li X, Liu Y, Alvarez BV, Casey JR, Fliegel L. A novel carbonic anhydrase II binding site regulates NHE1 activity. Biochemistry 2006;45: 2414–2424. 18. Dahl NK, Jiang L, Chernova MN, Stuart-Tilley AK, Shmukler BE, Alper SL. Deficient HCO32 transport in an AE1 mutant with normal Cl2 transport can be rescued by carbonic anhydrase II presented on an adjacent AE1 protomer. J Biol Chem 2003;278:44949–44958. 19. Lu J, Daly CM, Parker MD, Gill HS, Piermarini PM, Pelletier MF, Boron WF. Effect of human carbonic anhydrase II on the activity of the human electrogenic Na/HCO3 cotransporter NBCe1-A in Xenopus oocytes. J Biol Chem 2006;281:19241–19250. 20. Sterling D, Reithmeier RAF, Casey JR. A transport metabolon: functional interaction of carbonic anhydrase II and chloride/ bicarbonate exchangers. J Biol Chem 2001;276:47886–47894. 21. Parks SK, Chiche J, Pouyssegur J. pH control mechanisms of tumor survival and growth. J Cell Physiol 2011;226:299–308. 22. Gatenby RA, Gillies RJ. A microenvironmental model of carcinogenesis. Nat Rev Cancer 2008;8:56–61. 23. Birner P, Jesch B, Friedrich J, Riegler M, Zacherl J, Hejna M, Wrba F, Schultheis A, Schoppmann SF. Carbonic anhydrase IX overexpression is associated with diminished prognosis in esophageal cancer and correlates with Her-2 expression. Ann Surg Oncol 2011;18:3330–3337. 24. Lou Y, McDonald PC, Oloumi A, Chia S, Ostlund C, Ahmadi A, Kyle A, Auf dem Keller U, Leung S, Huntsman D, et al. Targeting tumor hypoxia: suppression of breast tumor growth and metastasis by novel carbonic anhydrase IX inhibitors. Cancer Res 2011;71: 3364–3376. 25. Poulsen SA. Carbonic anhydrase inhibition as a cancer therapy: a review of patent literature, 2007–2009. Expert Opin Ther Pat 2010;20:795–806. 26. McDonald PC, Winum J-Y, Supuran CT, Dedhar S. Recent developments in targeting carbonic anhydrase IX for cancer therapeutics. Oncotarget 2012;3:84–97. 27. Zhou R, Huang W, Yao Y, Wang Y, Li Z, Shao B, Zhong J, Tang M, Liang S, Zhao X, et al. CA II, a potential biomarker by proteomic analysis, exerts significant inhibitory effect on the growth of colorectal cancer cells. Int J Oncol 2013;43:611–621.

192

28. Liu LC, Xu WT, Wu X, Zhao P, Lv YL, Chen L. Overexpression of carbonic anhydrase II and Ki-67 proteins in prognosis of gastrointestinal stromal tumors. World J Gastroenterol 2013;19: 2473–2480. 29. Liu CM, Lin YM, Yeh KT, Chen MK, Chang JH, Chen CJ, Chou MY, Yang SF, Chien MH. Expression of carbonic anhydrases I/II and the correlation to clinical aspects of oral squamous cell carcinoma analyzed using tissue microarray. J Oral Pathol Med 2012;41: 533–539. 30. Parkkila S, Rajaniemi H, Parkkila AK, Kivela J, Waheed A, Pastorekova S, Pastorek J, Sly WS. Carbonic anhydrase inhibitor suppresses invasion of renal cancer cells in vitro. Proc Natl Acad Sci USA 2000;97:2220–2224. 31. Takahashi M, Yang XJ, Sugimura J, Backdahl J, Tretiakova M, Qian CN, Gray SG, Knapp R, Anema J, Kahnoski R, et al. Molecular subclassification of kidney tumors and the discovery of new diagnostic markers. Oncogene 2003;22:6810–6818. 32. Haapasalo J, Nordfors K, Jarvel ¨ a¨ S, Bragge H, Rantala I, Parkkila AK, Haapasalo H, Parkkila S. Carbonic anhydrase II in the endothelium of glial tumors: a potential target for therapy. Neuro-oncol 2007;9:308–313. 33. Hynninen P, Parkkila S, Huhtala H, Pastorekova S, Pastorek J, Waheed A, Sly WS, Tomas E. Carbonic anhydrase isozymes II, IX, and XII in uterine tumors. APMIS 2012;120:117–129. 34. Supuran CT, Scozzafava A. Carbonic anhydrase inhibitors—Part 94. 1,3,4-thiadiazole-2-sulfonamidederivatives as antitumor agents? Eur J Med Chem 2000;35:867–874. 35. Yoshiura K, Nakaoka T, Nishishita T, Sato K, Yamamoto A, Shimada S, Saida T, Kawakami Y, Takahashi TA, Fukuda H, et al. Carbonic anhydrase II is a tumor vessel endothelium–associated antigen targeted by dendritic cell therapy. Clin Cancer Res 2005;11: 8201–8207. 36. Mokhtari RB, Kumar S, Islam SS, Yazdanpanah M, Adeli K, Cutz E, Yeger H. Combination of carbonic anhydrase inhibitor, acetazolamide, and sulforaphane, reduces the viability and growth of bronchial carcinoid cell lines. BMC Cancer 2013;13:378–386. 37. Pan J, Copland I, Post M, Yeger H, Cutz E. Mechanical stretch–induced serotonin release from pulmonary neuroendocrine cells: implications for lung development. Am J Physiol Lung Cell Mol Physiol 2006;290: L185–L193. 38. Soll C, Jang JH, Riener MO, Moritz W, Wild PJ, Graf R, Clavien PA. 5-HT promotes tumor growth in human hepatocellular cancer. Hepatology 2010;51:244–254. 39. Drozdov I, Kidd M, Gustafsson BI, Svejda B, Joseph R, Pfragner R, Modlin IM. Autoregulatory effects of serotonin on proliferation and signaling pathways in lung and small intestine neuroendocrine tumor cell lines. Cancer 2009;115:4934–4945. 40. Nocito A, Dahm F, Jochum W, Jang JH, Georgiev P, Bader M, Graf R, Clavien PA. Serotonin regulates macrophage-mediated angiogenesis in a mouse model of colon cancer allografts. Cancer Res 2008;68:5152–5158. 41. Huber V, De Milito A, Harguindey S, Reshkin SJ, Wahl ML, Rauch C, Chiesi A, Pouyssegur ´ J, Gatenby RA, Rivoltini L, et al. Proton dynamics in cancer. J Transl Med 2010;8:57–60.
