Tumour biology: Herceptin acts as an anti-angiogenic cocktail
2
A lthough fluorine in the form of fluo- ride minerals is the most abundant halogen in the Earth’s crust, only 12 naturally occurring organofluorine com- pounds have so far been found 1 , and how these are biosynthesized remains a mystery 2 . Here we describe an enzymatic reaction that occurs in the bacterium Streptomyces cattleya and which catalyses the conversion of fluoride ion and S-adenosylmethionine (SAM) to 58-fluoro-58-deoxyfluoroadeno- sine (58-FDA). To our knowledge, this is the first fluorinase enzyme to be identified, a discovery that opens up a new biotechno- logical opportunity for the preparation of organofluorine compounds. The rarity of natural fluorinated prod- ucts contrasts with the identification of about 3,500 naturally occurring halogen- ated compounds 3 . The available fluoride is largely insoluble — for example, sea water contains 1.3 p.p.m. fluoride and 19,000 p.p.m. chloride, which may help to explain why fluorine’s biochemistry has hardly evolved. The toxin fluoroacetate is the most ubiq- uitious of the small class of organofluorine compounds and has been identified in more than 40 plant species from all of the continents apart from Antarctica 4 , but its biosynthetic fluorination pathway has not been clearly defined 2,5 . Fluoroacetate is also produced by the bacterium S. cattleya 6 when it is grown in culture medium sup- plemented with fluoride ions. We investigated this process of enzymatic fluorination by incubating a partially puri- fied protein extract from S. cattleya with fluoride ions and SAM, and monitored the reaction by using high-pressure liquid chromatography (HPLC). The initial prod- uct of fluorination in S. cattleya was shown to be 58-FDA by reference to a synthetic standard (Fig. 1 legend); the biotransformed compound and standard 58-FDA co-eluted on HPLC and (after derivatization) had identical molecular masses and 19 F nuclear magnetic resonsance (NMR) spectra. Our crude cell-free protein preparation from S. cattleya cells was able to mediate the biotransformation of SAM and fluoride ion all the way to fluoroacetate (Fig. 1), indicating that this organism also contains the necessary enzyme activities to convert 58-FDA to fluoroacetate. We incubated our synthetic 58-FDA with the crude protein extract of S. cattleya and directly monitored its bioconversion to fluoro- acetate by 19 F-NMR analysis and HPLC. The fluorination reaction seems to involve a nucleophilic attack by fluoride ion at the C-58 carbon of SAM, generating 58-FDA and concomitantly displacing L-methion- ine. The mechanism by which 58-FDA is metabolized to fluoroacetate remains to be established (Fig. 1), although fluoro- acetaldehyde 7,8 may be the immediate pre- cursor of fluoroacetate. David O’Hagan*, Christoph Schaffrath*, Steven L. Cobb*, John T. G. Hamilton†, Cormac D. Murphy* *School of Chemistry and Centre for Biomolecular Sciences, University of St Andrews, North Haugh, St Andrews, Fife KY16 9ST, UK e-mail: [email protected] †Microbial Biochemistry Section, Department of Food Science, The Queen’s University of Belfast, Newforge Lane, Belfast BT9 SPX, UK 1. O’Hagan, D. & Harper, D. B. Nat. Prod. Rep. 11, 123–133 (1994). 2. O’Hagan, D. & Harper, D. B. in Asymmetric Fluoro-organic Chemistry (ed. Ramachandran, P. V.) 210–225 (Am. Chem. Soc. Symp. Ser. 746, Washington DC, 1999). 3. Gribble, G. W. Chem. Soc. Rev. 28, 335–346 (1999). 4. Hall, R. J. New Phytol. 71, 855–871 (1972). 5. Peters, R. A. & Shorthouse, M. A. Nature 216, 80–81 (1967). 6. Sanada, M. et al. J. Antibiot. 39, 259–265 (1986). 7. Moss, S. J. Chem. Commun. 2281–2282 (2000). 8. Murphy, C. D., Moss, J. S. & O’Hagan, D. Appl. Environ. Microbiol. 67, 4919–4921 (2001). 9. Gani, D. & Johnson, A. W. J. Chem. Soc. Perkin Trans. 1, 1197–1204 (1982). Competing financial interests: declared none. brief communications NATURE | VOL 416 | 21 MARCH 2002 | www.nature.com 279 Biosynthesis of an organofluorine molecule A fluorinase enzyme has been discovered that catalyses carbon–fluorine bond formation. Figure 1 The fluorinase enzyme from Streptomyces cattleya mediates the conversion of S-adenosyl-L-methionine (SAM) to 58-fluoro-58-deoxyadenosine (58-FDA). NAD & , nicotine adenine- dinucleotide. A cell-free protein extract prepared by sonication of washed S. cattleya cells at 4 °C was precipitated with ammoni- um sulphate (45–60% saturation) and the precipitate dissolved in 50 mM Tris buffer, pH 7.8, before desalting on a HiTrap column (Pharmacia) and gradient elution from a 15 Q anion- exchange column (Pharmacia) with 50 mM Tris, pH 7.8, containing 0–400 mM KCl. Eluted fractions were incubated (960 ml) with SAM (0.4 mM) and KF (10 mM) for 16 h at 26 °C and assayed by high-pressure liquid chromatography (HPLC). 58-FDA was identified by gas chromatography with mass spectroscopy (GC-MS), HPLC and 19 F-NMR by comparison with a synthetic standard. Synthetic 58-FDA, prepared by treatment of 28-O,38-O-isopropylidene-58-O-p-tosyladenosine with tetra- butylammonium fluoride 9 and then with dilute sulphuric acid, yielded fluoroacetate when incubated with S. cattleya crude protein extract, identified by 19 F-NMR comparison to a reference sample. Details of spectroscopic characterization of standards are available from the authors. N N N N NH 2 O OH OH + H 3 N O O – + S–Me S-adenosylmethionine N N N N O F F – Fluorinase 5'-FDA O H F O F Fluoroacetaldehyde Fluoroacetate NAD + OH OH NH 2 OH Tumour biology Herceptin acts as an anti-angiogenic cocktail M alignant tumours secrete factors that enable them to commandeer their own blood supply (angiogenesis), and blocking the action of these factors can inhibit tumour growth. But because tumours may become resistant to treat- ments that target individual angiogenic factors by switching over to other angio- genic molecules 1,2 , a cocktail of multiple anti-angiogenic agents should be more effective. Here we show that herceptin 3 , a monoclonal antibody against the cell- surface receptor HER2 (for human epidermal growth factor receptor-2; ref. 4), induces normalization and regression of the vascu- lature in an experimental human breast tumour that overexpresses HER2 in mice, and that it works by modulating the effects of different pro- and anti-angiogenic factors. As a single agent that acts against multiple targets, herceptin, or drugs like it, may offer a simple alternative to combina- tion anti-angiogenic treatments. We found that herceptin treatment significantly reduces the diameter and volume, but not the length, of tumour blood vessels compared with those in tumours treated with a control antibody (Fig. 1a, b); vascular permeability is also significantly reduced (8.855.7 and 1.651.1210 17 cm s 11 at day 15 in the control and herceptin groups, respectively). Blood vessels in the herceptin-treated tumours thus more closely resemble a normal phenotype. Furthermore, tumour growth is slowed (Fig. 1b) and animal sur- vival is significantly extended (Table 1) in response to herceptin treatment. These vas- cular effects are not dependent on tumour size, as shown by their persistence until the end of the experiment when tumours were largest (6 mm in diameter; Table 1). To investigate how herceptin achieves this anti-angiogenic effect, we investigated the expression of 23 angiogenesis-related genes by using a gene array. Expression of the pro-angiogenic factors VEGF © 2002 Macmillan Magazines Ltd
Transcript of Tumour biology: Herceptin acts as an anti-angiogenic cocktail
A lthough fluorine in the form of fluo-ride minerals is the most abundanthalogen in the Earth’s crust, only 12
naturally occurring organofluorine com-pounds have so far been found1, and howthese are biosynthesized remains a mystery2.Here we describe an enzymatic reactionthat occurs in the bacterium Streptomycescattleya and which catalyses the conversionof fluoride ion and S-adenosylmethionine(SAM) to 58-fluoro-58-deoxyfluoroadeno-sine (58-FDA). To our knowledge, this is thefirst fluorinase enzyme to be identified, adiscovery that opens up a new biotechno-logical opportunity for the preparation oforganofluorine compounds.
The rarity of natural fluorinated prod-ucts contrasts with the identification ofabout 3,500 naturally occurring halogen-ated compounds3. The available fluoride is largely insoluble — for example, seawater contains 1.3 p.p.m. fluoride and19,000 p.p.m. chloride, which may help toexplain why fluorine’s biochemistry hashardly evolved.
The toxin fluoroacetate is the most ubiq-
uitious of the small class of organofluorinecompounds and has been identified inmore than 40 plant species from all of thecontinents apart from Antarctica4, but itsbiosynthetic fluorination pathway has notbeen clearly defined2,5. Fluoroacetate is alsoproduced by the bacterium S. cattleya6
when it is grown in culture medium sup-plemented with fluoride ions.
We investigated this process of enzymaticfluorination by incubating a partially puri-fied protein extract from S. cattleya with fluoride ions and SAM, and monitored thereaction by using high-pressure liquid chromatography (HPLC). The initial prod-uct of fluorination in S. cattleya was shownto be 58-FDA by reference to a synthetic standard (Fig. 1 legend); the biotransformedcompound and standard 58-FDA co-elutedon HPLC and (after derivatization) hadidentical molecular masses and 19F nuclearmagnetic resonsance (NMR) spectra.
Our crude cell-free protein preparationfrom S. cattleya cells was able to mediatethe biotransformation of SAM and fluorideion all the way to fluoroacetate (Fig. 1),indicating that this organism also containsthe necessary enzyme activities to convert58-FDA to fluoroacetate. We incubated our synthetic 58-FDA with the crude protein extract of S. cattleya and directlymonitored its bioconversion to fluoro-
acetate by 19F-NMR analysis and HPLC.The fluorination reaction seems to involvea nucleophilic attack by fluoride ion at theC-58 carbon of SAM, generating 58-FDAand concomitantly displacing L-methion-ine. The mechanism by which 58-FDA ismetabolized to fluoroacetate remains to be established (Fig. 1), although fluoro-acetaldehyde7,8 may be the immediate pre-cursor of fluoroacetate. David O’Hagan*, Christoph Schaffrath*,Steven L. Cobb*, John T. G. Hamilton†,Cormac D. Murphy**School of Chemistry and Centre for BiomolecularSciences, University of St Andrews, North Haugh, St Andrews, Fife KY16 9ST, UK e-mail: [email protected]†Microbial Biochemistry Section, Department ofFood Science, The Queen’s University of Belfast,Newforge Lane, Belfast BT9 SPX, UK
1. O’Hagan, D. & Harper, D. B. Nat. Prod. Rep. 11,123–133 (1994).
2. O’Hagan, D. & Harper, D. B. in Asymmetric Fluoro-organic
Chemistry (ed. Ramachandran, P. V.) 210–225
(Am. Chem. Soc. Symp. Ser. 746, Washington DC, 1999).
3. Gribble, G. W. Chem. Soc. Rev. 28, 335–346 (1999).
4. Hall, R. J. New Phytol. 71, 855–871 (1972).
5. Peters, R. A. & Shorthouse, M. A. Nature 216, 80–81 (1967).
6. Sanada, M. et al. J. Antibiot. 39, 259–265 (1986).
7. Moss, S. J. Chem. Commun. 2281–2282 (2000).
8. Murphy, C. D., Moss, J. S. & O’Hagan, D. Appl. Environ.
Microbiol. 67, 4919–4921 (2001).
9. Gani, D. & Johnson, A. W. J. Chem. Soc. Perkin Trans. 1,1197–1204 (1982).
Competing financial interests: declared none.
brief communications
NATURE | VOL 416 | 21 MARCH 2002 | www.nature.com 279
Biosynthesis of an organofluorine moleculeA fluorinase enzyme has been discovered that catalyses carbon–fluorine bond formation.
Figure 1 The fluorinase enzyme from Streptomyces cattleya
mediates the conversion of S-adenosyl-L-methionine (SAM) to
58-fluoro-58-deoxyadenosine (58-FDA). NAD&, nicotine adenine-
dinucleotide. A cell-free protein extract prepared by sonication of
washed S. cattleya cells at 4 °C was precipitated with ammoni-
um sulphate (45–60% saturation) and the precipitate dissolved
in 50 mM Tris buffer, pH 7.8, before desalting on a HiTrap
column (Pharmacia) and gradient elution from a 15 Q anion-
exchange column (Pharmacia) with 50 mM Tris, pH 7.8,
containing 0–400 mM KCl. Eluted fractions were incubated
(960 ml) with SAM (0.4 mM) and KF (10 mM) for 16 h at 26 °C
and assayed by high-pressure liquid chromatography (HPLC).
58-FDA was identified by gas chromatography with mass
spectroscopy (GC-MS), HPLC and 19F-NMR by comparison with a
synthetic standard. Synthetic 58-FDA, prepared by treatment of
28-O,38-O-isopropylidene-58-O-p-tosyladenosine with tetra-
butylammonium fluoride9 and then with dilute sulphuric acid,
yielded fluoroacetate when incubated with S. cattleya crude
protein extract, identified by 19F-NMR comparison to a reference
sample. Details of spectroscopic characterization of standards
are available from the authors.
N
NN
NNH2
O
OHOH
+H3N O
O–
+S–Me
S-adenosylmethionine
N
NN
N
OFF–
Fluorinase
5'-FDA
O
HF
O
F
FluoroacetaldehydeFluoroacetate
NAD+
OHOH
NH2
OH
Tumour biology
Herceptin acts as ananti-angiogenic cocktail
Malignant tumours secrete factors thatenable them to commandeer theirown blood supply (angiogenesis),
and blocking the action of these factors caninhibit tumour growth. But becausetumours may become resistant to treat-ments that target individual angiogenic factors by switching over to other angio-genic molecules1,2, a cocktail of multipleanti-angiogenic agents should be moreeffective. Here we show that herceptin3, amonoclonal antibody against the cell-surface receptor HER2 (for human epidermalgrowth factor receptor-2; ref. 4), inducesnormalization and regression of the vascu-lature in an experimental human breasttumour that overexpresses HER2 in mice,and that it works by modulating the effectsof different pro- and anti-angiogenic factors. As a single agent that acts againstmultiple targets, herceptin, or drugs like it,
may offer a simple alternative to combina-tion anti-angiogenic treatments.
We found that herceptin treatment significantly reduces the diameter and volume, but not the length, of tumourblood vessels compared with those intumours treated with a control antibody(Fig. 1a, b); vascular permeability is alsosignificantly reduced (8.855.7 and1.651.121017 cm s11 at day 15 in the control and herceptin groups, respectively).Blood vessels in the herceptin-treatedtumours thus more closely resemble a normal phenotype. Furthermore, tumourgrowth is slowed (Fig. 1b) and animal sur-vival is significantly extended (Table 1) inresponse to herceptin treatment. These vas-cular effects are not dependent on tumoursize, as shown by their persistence until theend of the experiment when tumours werelargest (6 mm in diameter; Table 1).
To investigate how herceptin achievesthis anti-angiogenic effect, we investigatedthe expression of 23 angiogenesis-relatedgenes by using a gene array. Expression of the pro-angiogenic factors VEGF
© 2002 Macmillan Magazines Ltd
other experimental models and also in aclinical setting. We anticipate that, as herceptin primarily affects tumour cellsdirectly and host cells indirectly, angiogenicfactors from host cells11 may eventually disrupt this more normal vasculature.Other anti-angiogenic partners (for exam-ple, an anti-VEGF agent) may therefore beneeded to achieve stable, long-term vesselregression. To identify optimal combina-tions, the contribution of host stromal cellsmust also be taken into account, togetherwith other angiogenesis-related factors andthe mechanisms that are responsible forcontrolling tumour–host interactions.
Other inhibitors of signal transductionthat have potential as cocktail treatmentsmight block different sets of angiogenicfactors. We should eventually be able toobtain angiogenic profiles of individual
tumours and patients, allowing the mostappropriate combination of signal-trans-duction inhibitors to be selected with aview to customizing cancer treatment. Yotaro Izumi, Lei Xu, Emmanuelle diTomaso, Dai Fukumura, Rakesh K. Jain Edwin L. Steele Laboratory, Department ofRadiation Oncology, Massachusetts GeneralHospital, Harvard Medical School, Boston,Massachusetts 02114, USAe-mail: [email protected]
1. Jain, R. K. & Carmeliet, P. F. Sci. Am. 285, 38–45 (2001).
2. Fidler, I. J. J. Natl Cancer Inst. 93, 1040–1041 (2001).
3. Slamon, D. J. et al. N. Engl. J. Med. 344, 783–792 (2001).
4. Yarden, Y. & Sliwkowski, M. X. Nature Rev. Mol. Cell Biol. 2,
127–137 (2001).
5. Kerbel, R. S., Viloria-Petit, A., Klement, G. & Rak, J.
Eur. J. Cancer 36, 1248–1257 (2000).
6. Allgayer, H. et al. J. Clin. Oncol. 18, 2201–2209 (2000).
7. Nicolini, G., Miloso, M., Moroni, M. C., Beguinot, L. &
Scotto, L. J. Biol. Chem. 271, 30290–30296 (1996).
8. Thurston, G. et al. Science 286, 2511–2514 (1999).
9. Bornstein, P. J. Clin. Invest. 107, 929–934 (2001).
10. Jain, R. K. Nature Med. 7, 987–989 (2001).
11.Fukumura, D. et al. Cell 94, 715–725 (1998).
12. Jain, R. K., Munn, L. L. & Fukumura, D. Nature Rev. Cancer
(in the press).
Competing financial interests: declared (see online version).
280 NATURE | VOL 416 | 21 MARCH 2002 | www.nature.com
(vascular-endothelial growth factor), TGF-a(transforming growth factor-a), Ang-1(angiopoietin-1) and PAI-1 (plasminogen-activator inhibitor-1) were all reduced,whereas expression of the anti-angiogenicfactor TSP-1 (thrombospondin-1) wasincreased in herceptin-treated tumours relative to control-treated tumours in vivo(Table 1). These expression profiles wereconfirmed by northern-blot analysis.
HER2 signalling is known to control theexpression of VEGF (ref. 5) and PAI-1 (ref.6). HER2 may also affect TGF-a transcriptsthrough interaction with HER1 (ref. 7) andmay mediate TSP-1 expression throughRas-like pathways5. To our knowledge, nocorrelation has been reported betweenHER2 and Ang-1. The vascular effects ofherceptin reported here are consistent withan increase in vessel diameter by Ang-1(ref. 8) and a reduction in vessel diameterby TSP-1 (ref. 9). Vascular permeability ispossibly affected by TSP-1 through recruit-ment of perivascular cells9, an effect thatwould be balanced by the reduction inAng-1 (ref. 8).
Surprisingly, VEGF expression wasreduced by herceptin treatment in vitrobut not in vivo (Table 1), suggesting thathost cells may produce compensatoryVEGF — an idea that is supported by our immunohistochemical observation ofVEGF in host stromal cells as well as intumour cells (data not shown). Culturedtumour cells also expressed smalleramounts of Ang-1 and PAI-1 and moreTSP-1 in vitro than tumour tissues did invivo, with the change in PAI-1 expressionbeing greater in vivo than in vitro. Herceptin treatment may affect PAI-1from the host by modulation of tumour-derived TGF-a. These results underline theimportance of tumour–host interactionsin the production of pro- and anti-angiogenic factors.
As the tumour vascular network afterherceptin treatment becomes more efficientand more closely resembles normal net-works, it might be possible to deliver drugsto previously inaccessible regions10. Howev-er, our findings need to be confirmed in
brief communications
Figure 1 Effects of herceptin on MDA-MB-361HK tumour growth, angiogenesis and gene expression. Tumours were grown in cranial
windows in severe combined immunodeficient (SCID) mice and intravital microscopy was carried out as described12. Established tumours
(about 2 mm at treatment day 0) were treated with herceptin or control immunoglobulin G (Genentech), at an intraperitoneal dose of
30 mg kg11 every 3 days. a, Vasculature of control (top) and herceptin-treated (bottom) tumours on treatment day 15. Scale bar, 100 mm.
b, Changes in vessel diameter and tumour size during treatment with herceptin (open circles) or control antibody (filled circles)
(mean 5s.d.; n46; asterisk denotes P*0.05).
* * * *
****
40
30
20
10
15
10
5
00 5 10 15
Ves
sel d
iam
eter
(µm
)Tu
mou
r si
ze (m
m2 )
Treatment days
a b
Table 1 Effects of herceptin on an experimental tumour
Vessel diameter Vessel density Vessel volume Permeability Survival
(mm) (cm cm12) (mm3 mm12) (cm s1121017) (days)
Control 40.456.7 144511 25.456.3 8.053.4 2053
Herceptin 14.254.1* 181570 3.852.3* 2.751.4* 33512*
Production of angiogenic factors
VEGF TGF-a Ang-1 PAI-1 TSP-1
Gene array 0.5 0.5 0.6 0.5 4.2
In vivo 0.9 0.5 0.7 0.2 5.3
In vitro 0.3 0.5 0.5 0.7 4.1
The physiological properties of 6-mm tumours (end point) are shown, together with the molecular profiles of angiogenesis-related factors produced by thesetumours. MDA-MB-361HK cells were treated in vitro with 50 mg ml11 herceptin or control human immunoglobulin G for 72 h. Values in the upper section ofthe table are means5s.d. (n46). In the lower section, messenger RNA abundance for each different factor is shown, normalized to b-actin and expressedas a ratio for herceptin-treated to control-treated tumours. Results were quantified by densitometry.*P*0.05 (Mann–Whitney U-test). Further details are available from the authors.
brief communications is intended to provide a forumfor both brief, topical reports of general scientific interestand technical discussion of recently published material ofparticular interest to non-specialist readers. Priority willbe given to contributions that have fewer than 500words, 10 references and only one figure. Detailed guidelines are available on Nature’s website(www.nature.com) or on request [email protected]
© 2002 Macmillan Magazines Ltd
![HERCEPTIN Herceptin [trastuzumab] Foundation of care in women with HER2-positive breast cancer Update 2010 Herceptin Small Tumor Efficacy is consistent.](https://static.fdocuments.us/doc/165x107/56649d2d5503460f94a04a57/herceptin-herceptin-trastuzumab-foundation-of-care-in-women-with-her2-positive.jpg)
