Nature - USA (2020-01-02)

Nature | Vol 577 | 2 January 2020 | 117

but more diffusely cytoplasmic in inefficient metastasizers (Extended
Data Fig. 2i).
The expression of MCT1 and CD147 (a co-chaperone of MCT1^26 ) did
not differ between primary and metastatic tumours (Extended Data
Fig. 3a–g), consistent with a previous study^27. We did not detect MCT2 in
any of the melanomas we studied (Fig. 2b). MCT4 was expressed (Fig. 2c;
see Extended Data Fig. 2b for quantification), but did not consistently
differ between primary and metastatic tumours (Extended Data Fig. 3b).

MCT1 is required during metastasis

To test whether MCT1 mediates lactate uptake by melanoma cells, we
transplanted efficiently metastasizing melanomas from three patients
subcutaneously into NSG mice, and then treated half of the mice for
7 days with the selective MCT1 inhibitor AZD3965 (30 mg kg−1 day−1),
which does not have activity against MCT4^28. We infused [U-^13 C]lactate
and measured the fractional enrichment in lactate relative to 3PG in
the tumours. In all three melanomas, AZD3965 treatment significantly
reduced lactate labelling, to the point that lactate and 3PG were equiva-
lently labelled, consistent with the labelled lactate arising from gly-
colysis rather than lactate uptake (Fig. 2f). Therefore, MCT1 mediates
lactate uptake in efficient metastasizers.
AZD3965 treatment did not significantly alter the levels of MCT1
(Extended Data Fig. 3h, i), CD147 (Extended Data Fig. 3j, k), β 1 integrin
(Extended Data Fig. 3n, o) or CD98 (Extended Data Fig. 3l, m) on the
surface of melanoma cells. In addition, AZD3965 treatment did not

significantly alter the levels of IKKα (Extended Data Fig. 3p–r) or IKKβ

(Extended Data Fig. 3s–u), or the epithelial–mesenchymal transition

markers E-cadherin (Extended Data Fig. 4a), N-cadherin (Extended

Data Fig. 4b) or vimentin (Extended Data Fig. 4c).

To test whether MCT1 inhibition affected primary tumour growth or

metastasis, we subcutaneously transplanted efficiently metastasizing

melanoma cells from three patients into NSG mice. Once tumours

were palpable, we treated every other day with AZD3965^29. AZD3965

had little effect on the growth of subcutaneous tumours (Fig. 2g) but

substantially reduced the frequency of circulating melanoma cells in

the blood (Fig. 2h), and metastatic disease burden in the same mice

(Fig. 2i, Extended Data Fig. 5).

We also infected melanoma cells from three patients with scram-

bled control short hairpin RNA (shRNA) or with shRNAs against MCT1

(also known as SLC16A1) (Extended Data Fig. 6a, b; these shRNAs did

not affect MCT4 expression) and then transplanted the cells subcu-

taneously into NSG mice. MCT1 knockdown had little effect on the

growth of the subcutaneous tumours (Extended Data Fig. 6c), but

significantly reduced the frequency of circulating melanoma cells

in the blood (Extended Data Fig. 6d), and metastatic disease burden

in all three melanomas (Extended Data Fig. 6e). The overexpression

of an shRNA-insensitive MCT1 cDNA (Extended Data Fig. 6f ) rescued

these effects (Extended Data Fig. 6h) without affecting subcutaneous

tumour growth (Extended Data Fig. 6g).

MCT1 overexpression in inefficiently metastasizing melanoma cells

significantly increased metastatic burden in vivo without affecting

M405 diameter (cm)

0.0

0.5

1.0

1.5

2.0

2.5

0.0

0.5

1.0

1.5

2.0

2.5

0.0

0.5

1.0

1.5

2.0

2.5

M481 diameter (cm)

UT10 diameter (cm)

50 75 100 125

Days

60 80 100 120

Days

40 60 80

Days

0.00

0.02

0.04

0.06

Melanoma

cells in blood (%)

0

1

2

3

Metastatic burden(relative total flux)

0.00

0.02

0.04

0.06

Melanoma

cells in blood (%)

0

1

2

3

Metastatic burden(relative total flux)

ghi

0

1

2

3

Metastatic burden(relative total flux)

DMSOAZD

0.0

0.5

1.0

1.5

Melanoma

cells in blood (%)

DMSOAZD

DMSO

AZD

DMSO

AZD

DMSOAZD DMSOAZD

3.0

a

WT KO 123123

HCC15

MCT1

Actin

Efficient Inefficient

NS

0.08

0.10

0.0

1.0

2.0

3.0

YUMM 5.2 diameter (cm)

Days

20 40 60

Metastatic burden(relative total flux)

DM

SOAZD

4

0.0

0.5

1.0

1.5

2.0

YUMM 1.7 diameter (cm)

Days

10 15 20

Metastatic burden(relative total flux)

DMSOAZD

jk

0.0

1.0

2.0

3.0

YUMM 3.3 diameter (cm)

Days

30 40 50

DMSO, n = 5

AZD3965,n = 5

Metastatic burden(relative total flux)

DMSOAZD

123123

Efficient Inefficient

4

MCT2

Actin

MCF7

MCT4

Actin

b

c

DMSO, n = 8

AZD3965,n = 10

DMSO, n = 11

AZD3965,n = 11

DMSO, n = 10

AZD3965,

n = 10 20

18

DMSO, n = 10

AZD3965,

n = 10

DMSO, n = 10

AZD3965,n = 10

WT KO 123 123

HCC15Efficient Inefficient

4

MCT1-Alexa Fluor647

M528 IsotypeM597

Isotypecontrol control

Inefficient

d

M481 UT10

Isotypecontrol Isotypecontrol

e

Isotypecontrol

IsotypecontrolM405

M610

Efficient

MCT1-Alexa Fluor647

MCT1 MCT1 MCT1

MCT1 MCT1 MCT1

f

Fractional enrichmentlactate m+3/3PG m+3

M405 M481 UT10

0.0

1.0

2.0

3.0

444546

DMSO

P = 0.010AZD3965

P = 0.0006

0

2

4

0

2

4

6

0

1

3

5

P = 0.006

0.0001P <

0.0002P =

0.0495P =

0.0001P <

P = 0.0005

P = 0.052

P = 0.007

P = 0.0006

P = 0.022

P = 0.007

0.0001P <

0.0001P <

18

20

15

17

15

17

10

10 10

10

13

15

13

11

16

16

Fig. 2 | MCT1 inhibition selectively impairs metastasis in human and mouse
melanomas. a–c, Western blot analysis of MCT1 (a), MCT2 (b) and MCT4 (c) in
three efficiently (M405, M481 and UT10) and four inefficiently (M498, M528,
M597 and M610) metastasizing xenografted melanomas. Wild-type (WT)
HCC15 cells were positive controls for MCT1 and MCT4; MCT1- and MCT4-
deficient (KO) HCC15 cells were negative controls for MCT1 and MCT4,
respectively. MCF7 cells were a positive control for MCT2. The data are
representative of four (a) or two (b, c) experiments. d, e, Flow cytometric
analysis of MCT1 surface expression in inefficiently (d) and efficiently (e)
metastasizing melanomas. f, Enrichment of lactate m + 3 normalized to 3PG
m + 3 in xenografted tumours after treatment with the MCT1 inhibitor AZD3 965
or DMSO control and infusion of [U-^13 C]lactate (two experiments per

melanoma). The number of mice per treatment is indicated. g–i, Growth of

subcutaneous tumours (g) in mice treated with AZD3965 (AZD) or DMSO

control; the frequency of circulating melanoma cells in the blood (h); and

metastatic disease burden based on bioluminescence imaging (i). Data in h and

i reflect one (UT10) or two experiments per melanoma, but only one

representative experiment per melanoma is shown in g. j, k, Growth of

subcutaneous tumours (j) and metastatic disease burden at end point by

bioluminescence imaging (k) in mice transplanted with YUMM1.7, YUMM3.3 or

YUMM5.2 mouse melanomas and treated with AZD3965 or DMSO control (two

experiments per melanoma). Data are mean ± s.d. Statistical significance was

assessed using t-tests (f), nparLD (g), mixed-effects analysis (j) or

Mann–Whitney tests (h, i, k). NS, not significant.