Commentary
`
`Can engineered bacteria help control cancer?
`
`Rakesh K. Jain* and Neil S. Forbes
`
`Edwin L. Steele Laboratory, Department of Radiation Oncology, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114
`
`Hypoxia and anoxia are pathophysio-
`
`logic characteristics of most solid tu-
`mors (1, 2). For nearly 150 years, non-
`pathogenic, anaerobic bacteria that
`preferentially localize and proliferate in
`the hypoxic regions of tumors have been
`investigated as treatments for experimen-
`tal and human tumors with mixed success
`(Table 1). In recent years, there has been
`a renewed interest in using these bacteria
`as innovative delivery vehicles for gene
`therapy (Table 1). Now, as described in
`this issue of PNAS, Vogelstein and co-
`workers (11) have created a new strain of
`anaerobic bacteria, devoid of its toxic
`genes, that leads to dramatic and pro-
`longed regression of subcutaneous tumors
`when systematically administered with
`conventional drugs. This strategy, re-
`ferred to as combination bacteriolytic
`therapy (COBALT), adds a new weapon
`in the war against cancer. However, there
`are still obstacles that need to be over-
`come before it can be used safely in the
`clinic.
`In tumors, blood vessels are structurally
`and functionally abnormal, resulting in
`temporally and spatially heterogeneous
`blood flow (19, 20). This heterogeneity
`hinders the delivery
`of blood-borne ther-
`apeutics to all cancer
`cells and leads to
`and兾or
`acutely
`chronically hypoxic
`and acidic regions in
`tumors
`(Fig.
`1).
`These conditions re-
`duce the effective-
`ness of radiation and some chemothera-
`peutic agents and select for cancer cells
`that are more aggressive, metastatic, and
`resistant to various therapies (2, 21).
`Ironically, a tumor’s metabolically com-
`promised microenvironment provides a
`haven for a number of anaerobic bacteria.
`And indeed, over the past 50 years, several
`strains of facultative and obligate anaer-
`obic bacteria have been shown to localize
`and cause lysis in transplanted tumors in
`animals (Table 1). These initial animal
`studies were so encouraging that clinical
`trials using Clostridium began in the 1960s
`(8). Unfortunately, the results were not as
`impressive as anticipated and the trials
`were discontinued.
`
`Ironically, a tumor’s metabolically
`compromised microenvironment
`provides a haven for a number of
`anaerobic bacteria.
`
`So why is there a resurgence of interest
`in using bacteria to treat solid tumors? To
`answer this question we need to examine
`the criteria for an ideal anticancer
`bacterium.
`They should be: (i) nontoxic to the
`host; (ii) only able to replicate within the
`tumor; (iii) motile and able to disperse
`evenly throughout a tumor (including
`hypoxic and necrotic regions); (iv) slowly
`and completely eliminated from the host;
`(v) nonimmunogenic; and (vi) able to
`cause lysis of tumor cells by direct com-
`petition for nutrients, localized produc-
`tion of cytotoxins, or production of ther-
`apeutic amplifiers.
`In the last decade, significant progress
`has been made on each of these fronts.
`Multiple approaches have been used to
`remove the toxin genes of bacteria (16,
`17). For instance, Dang et al. (11) used
`heat shock to eliminate the lethal toxin
`genes from Clostridium novyi,
`located
`within a phage episome. Modern molec-
`ular approaches might be used once ge-
`nome sequences of various strains of bac-
`teria become available (22, 23). Of course,
`the use of naturally nonpathogenic bacte-
`ria (e.g., Clostridium oncolyticum) might
`avoid the toxicity
`problem
`alto-
`gether. Addition-
`ally, techniques de-
`veloped to transfer
`genetic material
`into bacteria other
`than Escherichia
`coli,
`for example
`the anaerobic bac-
`teria Clostridium acetobutylicum (24) and
`Bifidobacterium longum (25), have the po-
`tential to modulate the toxicity, motility,
`and protein expression of therapeutic
`bacteria.
`Currently there are no rapid, reliable,
`and inexpensive methods to screen for an
`ideal bacterium. Dang et al. (11) screened
`26 strains of bacteria for their ability to
`spread evenly throughout poorly vascular-
`ized regions of tumors. The selected bac-
`teria were seen growing throughout the
`enlarged necrotic regions of tumors after
`systemic injection of spores. Apparently,
`the bacteria were destroying the viable
`cells at the interface of the necrotic re-
`gion, and using the degradation products
`
`as nutrients. However, this treatment did
`not eradicate the tumor completely, leav-
`ing a ring of viable cells at the tumor
`periphery. To kill cells in the viable ring,
`Dang et al. chose to combine the bacte-
`riolytic therapy with low molecular weight
`conventional chemotherapy (mitomycin C
`and cytoxan). Their rationale was that the
`bacteria would lyse the tumors from the
`inside out, and low molecular weight che-
`motherapeutic agents would attack cancer
`cells in the well-perfused, non-necrotic
`region, a concept used since 1964 (7)
`(Table 1).
`To enhance the effect of chemothera-
`peutics (mitomycin C and cytoxan) and
`bacteria, Dang et al. used dolastatin (D-
`10), an antivascular agent. To our knowl-
`edge, this is the first time antivascular
`therapy has been combined with bacterio-
`lytic therapy. The benefit of this addition
`to COBALT, as described by Dang et al.,
`is that vascular stasis increases the extent
`of hypoxia thereby increasing the size of
`the region affected by C. novyi. It appears
`that this combination is the predominant
`reason for the effectiveness of COBALT.
`A problem with the low molecular weight
`chemotherapeutics is that they are rapidly
`cleared from perfused regions (i.e., the
`viable ring) (26). The additional benefit of
`including antivascular agents that lead to
`vascular shutdown is that they can trap
`extravasated molecules in tumors (27),
`thereby enhancing exposure to therapeu-
`tic agents in combination therapy.
`Indeed, COBALT therapy did produce
`impressive results. Dang et al. treated two
`different tumor lines grown subcutane-
`ously in mice and observed regression in
`most tumors and complete cure in a con-
`siderable proportion of mice that sur-
`vived. Whether similar cure rates can be
`achieved with COBALT in orthotopic
`and spontaneous tumors needs to be
`examined.
`Besides COBALT, there are several
`other strategies that amplify bacteriolytic
`therapy. One of these is to engineer bac-
`teria to produce inflammatory cytokines
`(e.g., tumor necrosis factor ␣) that in-
`
`See companion article on page 15155.
`
`*To whom reprint requests should be addressed. E-mail:
`jain@steele.mgh.harvard.edu.
`
`14748 –14750 兩 PNAS 兩 December 18, 2001 兩 vol. 98 兩 no. 26
`
`www.pnas.org兾cgi兾doi兾10.1073兾pnas.261606598
`
`Genome Ex. 1055
`Page 1 of 3
`
`

`

`COMMENTARY
`
`Nosignificantbenefitobserved
`
`Investigatebenefitofnonspecific
`
`immuno-stimulatingagents
`
`overotherorgans
`
`Accumulateintumors⬎1,000-fold
`
`AttenuatedSalmonellatargettumor
`
`8yrs
`
`30d
`
`Significantdelayintumorgrowth
`
`AttenuatetoxicityofSalmonellaand
`
`40d
`
`retaintumortargeting
`
`spontaneoustumors
`
`EngineeredB.longumalsotargets
`
`Testdeliverytospontaneoustumors
`
`Engineeredbacteriafoundonlyin
`
`Genedeliveryusingengineered
`
`tumors
`
`Bifidobacterium
`
`7d
`
`7d
`
`nontoxic
`
`B.bifidumlocalizestotumorsandis
`
`B.bifidumtoidentifytumors
`
`90d
`
`Tumorregression
`
`ofmice
`complicatedbydeathof⬃15–45%
`
`Completecurein⬃50%ofmice,
`
`intumorlysate
`
`Targetedimmunomodulation
`
`tumormasses
`bacteriolytictherapytoshrink
`anti-vasculartherapywith
`Usecombinationchemo-and
`
`convert5-FCto5-FU
`
`Cytosinedeaminaseactivitydetected
`
`Clostridiumdeliveryofenzymeto
`
`30d
`
`3mo
`
`Chemo-therapeutics
`antivascular
`Cytotoxicand
`
`Mouse
`
`Increasedgrowthdelay,eventual
`
`Heavymetalsincreasetumorlysis
`
`death
`
`differentClostridiumspecies
`
`Regressionbuteventualanimal
`
`Identificationofmosteffectiveof14
`
`Short-until
`
`Mouse
`
`regionsintumors
`C.tetanitohypoxicandnecrotic
`
`Rapiddeathoftumorbearingmice
`
`Localizationofobligatoryanaerobic
`
`survival
`
`withantitoxinandpenicillin
`
`Temporaryregressionandprolonged
`
`Diminishtoxicityofhistolyticum
`
`2d
`
`66d
`
`Mouse
`
`Results
`
`Strategy
`
`Length
`
`Combination
`
`Animal
`
`3mo
`
`⬍12d
`
`death
`
`E-39,MitomycinC
`5-FU,Tetramin,
`dextran
`Heavymetal-iron
`
`Hamster
`Mouse
`Rat
`
`tumorlysate
`
`Nitroreductaseactivitydetectedin
`
`death,lysiswithsurvival
`
`Threecases:failuretolyse,lysiswith
`Within3monthsallanimalsdied.
`significantlyincreasedregression.
`metastases.Allcombinations
`notaffectsmalltumorsor
`
`Clostridiumregressedtumorsbutdid
`
`animaldeath
`
`CB1954
`nitroreductasetoactivateprodrug
`
`Clostridiumdeliveryof
`
`Clostridium-inducedlysis
`
`13mo
`
`viablerim
`variouschemotherapeuticstokill
`UsecombinationofClostridiumand
`
`HumanMelphalan,5-FU
`Monkey
`Mouse
`
`1990Breastcancer
`
`18
`
`parvum
`
`Corynebacterium
`
`2000B16–F10melanoma
`
`17
`
`typhimurium
`
`anaerobe)
`(obligatory
`
`anaerobe)
`(facultative
`
`Mouse
`
`1999B16–F10melanoma
`
`16
`
`typhimurium
`
`Salmonella
`
`carcinoma
`mammary
`
`Rat
`
`2001DMBA-induced
`
`15
`
`Mouse
`
`2000B16–F10melanoma
`
`14
`
`Mouse
`
`1980Fibrosarcoma
`
`13
`
`longum
`
`longum
`
`bifidum
`
`Mouse
`
`1978Meth-Asarcoma
`
`12
`
`infantis
`
`Bifidobacterium
`
`anaerobe)
`(obligatory
`
`carcinoma
`HCT116colon
`2001B16melanoma,
`
`11
`
`novyi
`
`Rat
`
`2001Rhabdomyo-sarcoma
`
`10
`
`acetobutylicum
`
`Mouse
`
`Human
`
`1997EMT6
`
`1967
`
`adenocarcinoma
`renal
`
`1964Sarcoma,melanoma,
`
`1964Carcinoma,melanomaMouse,
`
`1964Ehrlichcarcinomas
`
`1955Carcinoma,hepatomaMouse
`
`1947Sarcoma
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`(acetobutylicum)
`
`beijerinckii
`
`butyricum
`
`tyrobutyricum
`pectinovorum
`butyricum
`acetobutylicum
`
`(M-55)
`
`butyricum
`
`(M-55)
`
`butyricum
`
`tetani
`
`anaerobe)
`(obligatory
`
`histolyticum
`
`Clostridium
`
`Model
`
`Ref.Year
`
`Species
`
`Organism
`
`Table1.Examplesofbacteriolytictherapyoftumorsinvivo
`
`Jain and Forbes
`
`PNAS 兩 December 18, 2001 兩 vol. 98 兩 no. 26 兩 14749
`
`Genome Ex. 1055
`Page 2 of 3
`
`

`

`Schematic of three microenvironmental regions in a centrally necrotic tumor. A spontaneous tumor may consist of many such necrotic foci. Decreasing
`Fig. 1.
`magnitude of various physiological parameters is indicated as ⫹⫹⫹, ⫹⫹, ⫹, ⫹兾⫺, and ⫺.
`
`crease the sensitivity of tumors to radia-
`tion therapy and兾or evoke a host immune
`response (28). Another approach is bac-
`teria-directed enzyme prodrug therapy
`(BDEPT), a variation of antibody-
`directed enzyme prodrug
`therapy
`(ADEPT). In this approach, targeting
`bacteria are engineered to produce en-
`zymes that can activate prodrugs within
`the tumor (9, 29). Another possibility is to
`place the genes of prodrug-activating en-
`zymes under the control of radiation-
`inducible promoters to provide spatial and
`temporal control, thus enabling selective
`killing of tumor cells while sparing normal
`cells (28, 30).
`So what are the potential problems with
`bacteriolytic therapy? First, there is the
`immediate problem encountered by Dang
`et al.: toxicity. Even after removing the
`toxin genes, COBALT therapy led to
`⬃15–45% mortality in mice. Whether this
`is caused by the so-called tumor lysis syn-
`drome (31) or the efflux of toxic bacterial
`products is not known. Identification of
`
`1. Helmlinger, G., Yuan, F., Dellian, M. & Jain, R. K. (1997)
`Nat. Med. 3, 177–182.
`2. Brown, J. M. & Giaccia, A. J. (1998) Cancer Res. 58,
`1408–1416.
`3. Parker, R. C., Plummer, H. C., Siebenmann, C. O. &
`Chapman, M. G. (1947) Proc. Soc. Exp. Biol. Med. 66,
`461–467.
`4. Malmgren, R. A. & Flanigan, C. C. (1955) Cancer Res. 15,
`473–478.
`5. Mo¨se, J. R. & Mo¨se, G. (1964) Cancer Res. 24, 212–216.
`6. Gericke, D. & Engelbart, K. (1964) Cancer Res. 24, 217–221.
`7. Thiele, E. H., Arison, R. N. & Boxer, G. E. (1964) Cancer
`Res. 24, 222–233.
`8. Carey, R. W., Holland, J. F., Whang, H. Y., Neter, E. &
`Bryant, B. (1967) Eur. J. Cancer 3, 37–46.
`9. Lemmon, M. J., van Zijl, P., Fox, M. E., Mauchline, M. L.,
`Giaccia, A. J., Minton, N. P. & Brown, J. M. (1997) Gene
`Ther. 4, 791–796.
`10. Theys, J., Landuyt, W., Nuyts, S., Van Mellaert, L., van
`Oosterom, A., Lambin, P. & Anne, J. (2001) Cancer Gene
`Ther. 8, 294–297.
`11. Dang, L. H., Bettegowda, C., Huso, D. L., Kinzler, K. W.
`& Vogelstein, B. (2001) Proc. Natl. Acad. Sci. USA 98,
`15155–15160. (First Published November 27, 2001;
`10.1073兾pnas.251543698)
`12. Kohwi, Y., Imai, K., Tamura, Z. & Hashimoto, Y. (1978)
`Gann 69, 613–618.
`
`the toxins released by rapidly lysing tu-
`mors or by large colonies of Clostridium
`contained within tumors is essential for
`alleviating the toxicity. Toxins expressed
`by the bacteria may be identified after
`complete sequencing of the respective ge-
`nomes. Well-known strategies then can be
`used to tackle specific toxins. On the other
`hand, alleviating the toxicity from low
`molecular weight byproducts of dying cells
`will require careful control of the rate of
`tumor lysis.
`Once the issues of systemic toxicity and
`incomplete tumor lysis are addressed,
`there are other potential pitfalls that may
`impede the success of COBALT therapy
`in the clinic. The most significant of these
`is acquired drug resistance, which lowers
`the effectiveness of the standard chemo-
`therapeutics used in COBALT after re-
`peated treatment. Even new drugs such as
`Gleevec are facing this age-old problem
`(32). However, antiangiogenic and anti-
`vascular agents may be less susceptible to
`this type of resistance (21, 33). Combined
`
`bacteriolytic antiangiogenesis therapy
`(COMBAT) may, thus, overcome or cir-
`cumvent the problem of drug resistance.
`A third and more difficult problem is
`that of treating small non-necrotic metas-
`tases of large primary tumors. The current
`strategy is to treat metastases as early as
`possible with conventional chemothera-
`peutics before the onset of physiological
`and兾or multidrug resistance. COBALT
`would require one to wait until the me-
`tastases develop hypoxic兾necrotic re-
`gions. Because metastasis is the major
`cause of mortality from cancer (34), we
`wonder whether it would be possible to
`engineer bacteria that can localize in small
`orthotopic tumors and their spontaneous
`metastases that do not contain large hy-
`poxic regions? Such bacteria would not
`only facilitate treatment of metastases but
`also their early detection by using molec-
`ular imaging techniques.
`
`We thank Drs. Brenda Fenton and Martin
`Brown for helpful discussions.
`
`13. Kimura, N. T., Taniguchi, S., Aoki, K. & Baba, T. (1980)
`Cancer Res. 40, 2061–2068.
`14. Yazawa, K., Fujimori, M., Amano, J., Kano, Y. & Tan-
`iguchi, S. (2000) Cancer Gene Ther. 7, 269–274.
`15. Yazawa, K., Fujimori, M., Nakamura, T., Sasaki, T.,
`Amano, J., Kano, Y. & Taniguchi, S. (2001) Breast Cancer
`Res. Treat. 66, 165–170.
`16. Low, K. B., Ittensohn, M., Le, T., Platt, J., Sodi, S., Amoss,
`M., Ash, O., Carmichael, E., Chakraborty, A., Fischer, J.,
`et al. (1999) Nat. Biotechnol. 17, 37–41.
`17. Clairmont, C., Lee, K. C., Pike, J., Ittensohn, M., Low,
`K. B., Pawelek, J., Bermudes, D., Brecher, S. M., Margi-
`tich, D., Turnier, J., et al. (2000) J. Infect. Dis. 181,
`1996–2002.
`18. Fisher, B., Brown, A., Wolmark, N., Fisher, E. R., Red-
`mond, C., Wickerham, D. L., Margolese, R., Dimitrov, N.,
`Pilch, Y., Glass, A., et al. (1990) Cancer 66, 220–227.
`19. Jain, R. K. (1988) Cancer Res. 48, 2641–2658.
`20. Jain, R. K. (1998) Nat. Med. 4, 655–657.
`21. Carmeliet, P. & Jain, R. K. (2000) Nature (London) 407,
`249–257.
`22. McClelland, M., Sanderson, K. E., Spieth, J., Clifton, S. W.,
`Latreille, P., Courtney, L., Porwollik, S., Ali, J., Dante, M.,
`Du, F., et al. (2001) Nature (London) 413, 852–856.
`23. Parkhill, J., Dougan, G., James, K. D., Thomson, N. R.,
`Pickard, D., Wain, J., Churcher, C., Mungall, K. L.,
`Bentley, S. D., Holden, M. T., et al. (2001) Nature (London)
`413, 848–852.
`
`24. Oultram, J. D., Peck, H., Brehm, J. K., Thompson, D. E.,
`Swinfield, T. J. & Minton, N. P. (1988) Mol. Gen. Genet.
`214, 177–179.
`25. Matsumura, H., Takeuchi, A. & Kano, Y. (1997) Biosci.
`Biotechnol. Biochem. 61, 1211–1212.
`26. Jain, R. K. & Baxter, L. T. (1988) Cancer Res. 48, 7022–
`7032.
`27. Pedley, R. B., Hill, S. A., Boxer, G. M., Flynn, A. A.,
`Boden, R., Watson, R., Dearling, J., Chaplin, D. J. &
`Begent, R. H. (2001) Cancer Res. 61, 4716–4722.
`28. Nuyts, S., Van Mellaert, L., Theys, J., Landuyt, W., Bos-
`mans, E., Anne, J. & Lambin, P. (2001) Gene Ther. 8,
`1197–1201.
`29. Fox, M. E., Lemmon, M. J., Mauchline, M. L., Davis, T. O.,
`Giaccia, A. J., Minton, N. P. & Brown, J. M. (1996) Gene
`Ther. 3, 173–178.
`30. Nuyts, S., Van Mellaert, L., Theys, J., Landuyt, W.,
`Lambin, P. & Anne, J. (2001) Radiat. Res. 155, 716–723.
`31. Altman, A. (2001) Semin. Oncol. 28, 3–8.
`32. McCormick, F. (2001) Nature (London) 412, 281–282.
`33. Folkman, J. (2001) in Harrison’s Textbook of Internal
`Medicine, eds. Braunwald, E., Fauci, A. S., Kasper, D. L.,
`Hauser, S. L., Longo, D. L. & Jameson, J. L. (McGraw–
`Hill, New York), pp. 517–530.
`34. Fidler, I. J., Singh, R. K., Yoneda, J., Kumar, R., Xu, L.,
`Dong, Z., Bielenberg, D. R., McCarty, M. & Ellis, L. M.
`(2000) Cancer J. Sci. Am. 6, S225–S236.
`
`14750 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.261606598
`
`Jain and Forbes
`
`Genome Ex. 1055
`Page 3 of 3
`
`