Cell line‐based assessment of BTK inhibitors

Abbreviations

BCR

B cell receptor

BTK

Bruton's tyrosine kinase

TK

Tyrosine kinase

TEL

ETV6 (Tel oncogene)

We read with great interest the accepted manuscript by Gui et al. (2019) on the novel kinase inhibitor XMU‐MP‐3 and its characterization by means of cell line models representing malignant B cells. Among these, engineered BaF3 cells expressing oncogenic TEL‐https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1948 fusions are of particular interest, since they permit appraisal of resistance mutations. In our letter, we wish to share considerations on the use of cell line models as tools for comparatively rating the potency and target selectivity of inhibitors along with functional roles of targeted kinases.

When N‐terminally fused to TEL, tyrosine kinases (TKs) are capable of transforming https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4994‐dependent BaF3 cells to factor‐independent growth that is highly sensitive to inhibition of the respective TKs. By comparison with IL‐3‐complemented parental BaF3 cells, their transformed derivatives enabled differential cytotoxicity screening of substance libraries (Melnick et al., 2006). Whereas this work opened a new dimension to the high‐throughput screening of compound libraries by using multiple engineered cell lines in parallel, Gui et al. (2019) applied their cell line panel to a single inhibitor, eliminating the screening aspect (Table 1). Of merit, they extended the range of available TEL fusions of TKs to BTK. The http://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=553 inhibitor https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=5678, then called BMS‐354825, was found by Melnick et al. (2006) to target https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=327 receptors, but only later to target BTK as well. However, one cannot emphasize enough that the transforming property of TEL‐BTK is not shared by BTK as such. Failing to mention the TEL fusion, the designation “BaF3‐BTK” in the paper at question is misleading. Owing to the high oncogenicity of TEL fusions, inhibitor potencies observed with transformed BaF3 cells closely resemble those determined on purified enzymes (Melnick et al., 2006). Conversely, investigation of the dependence of cell functions on intracellular BTK activity (Göckeritz et al., 2017) or of the transforming properties of BTK mutants (Wang et al., 2019) is possible only with unchanged, non‐activated BTK. In summary, BaF3‐TEL‐BTK cells, owing to the activating and transforming properties of the TEL‐BTK fusion, appear more appropriate as a substitute of biochemical assays of inhibitor potency than for assessing cell functions.

Table 1.

Tel‐ TKs that transformed BaF3 cells IL‐3‐independentgrowth and enabled differential cytotoxicity screening

https://www.guidetopharmacology.org/GRAC/ReceptorFamiliesForward?type=ENZYME&familyId=936			https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=304&familyType=ENZYME
Family/subfamily	Melnick et al. (2006)	Gui et al. (2019)	Family	Melnick et al. (2006)	Gui et al. (2019)
FAK		FAK	EGFR		EGFR
ABL	ABL1	ABL1a			ERBB2
		ABL2			(HER)
		(ARG)	ALK	ALK	ALK
TEC	BMX	BTK b	INSR	INSR	INSR
		BMX			IGF1R
SRC/FRK	FRK	FRK	CCK4		CCK4
SRC/SRM		BRK	TRK	TRKA
SRC/SRCA		SRC		TRKB
SRC/SRCB	BLK	HCK	MET	MET
	LCK	LCK		RON
	LYN	LYN	RET	RET
JAK	JAK2		FGFR	FGFR1	FGFR1
	JAK3s			FGFR2	FGFR2
	TYK				FGFR3
					FGFR4
			VEGFR		KDR
			PDGFR	FLT3	PDGFRA
				PDGFRB	PDGFRB
			TIE	TIE2	TIE1
			EPH	EPHB1	EPHA1
				EPHB2	EPHA3
				EPHB4	EPHA4

Note. Among 90 TKs of the human kinome, Melnick et al. (2006) and Gui et al. (2019) used partly overlapping sets of 30 TK targets each. Kinases that are unique to one of the two sets are printed in italics and underlined. The targets are arranged according to families of cytoplasmic and receptor TKs. The assignment to TK families is according to KinBase (http://www.kinase.com; accessed Sep. 23, 2019). Both sets are arbitrary and incomplete representations of the human kinome and even of the subgroup of TKs.

^aBCR‐ABL used here in place of TEL‐ABL fusion.

^bEssential for characterizing BTK inhibitors including XMU‐MP‐3 and dasatinib.

Gui et al. (2019) determined the effects of mutations at the binding pocket on the cellular potencies of BTK inhibitors in TEL‐BTK‐transformed BaF3 cells. The C481S mutation diminished the cellular potency of XMU‐MP‐3 approximately 100 times less than that of https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=6912, in line with the expectation that Cys‐481 is involved in the interaction of the BTK binding pocket only with irreversible, but not reversible, inhibitors. Diminished potency owing to mutation of the gatekeeper mutation at Thr‐474, as observed here for XMU‐MP‐3 and previously, for example, for dasatinib, is typical of reversible BTK inhibitors, unless the steric hindrance of inhibitor binding is avoided by particular molecular design. In cell line models, both mentioned resistance mutations served as chemical genetic tools to demonstrate involvement of BTK in cell functions like B cell receptor (BCR)‐stimulated chemokine secretion and CXCL12‐driven chemotaxis (Göckeritz et al., 2017). However, this approach is not applicable to TEL‐BTK‐transformed BaF3 cells, in which targeting of BTK by XMU‐MP‐3 was proved by differential cytotoxicity against BaF3‐TEL‐BTK versus parental BaF3 cells and not of wild type versus T474I mutation. With regard to resistance mutations, XMU‐MP‐3 showed expected features of a reversible BTK inhibitor.

The reversible BTK inhibitors dasatinib, https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=9299, and ARQ 531 showed approximately 2, 5, and 10 times stronger biochemical potencies against BTK than XMU‐MP‐3. The examination of its target selectivity in an arbitrary panel of recombinant BaF3 cells (Table 1) lacked reference substances but revealed that XMU‐MP‐3 resembles dasatinib in targeting BTK and ABL and differs in not inhibiting, for example, the Src family kinases http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=2060 and http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=2053. The assessment of XMU‐MP‐3 for cytotoxicity against cell lines and inhibition of anchorage‐dependent growth was not controlled for effects that are not due to inhibition of BTK activity, and the latter completely lacked reference substances. For these reasons, the cell line‐based characterization of XMU‐MP‐3 remains to be complemented.

This preclinical assessment of a reversible BTK inhibitor was entirely accomplished by means of cell lines, which even served as subcutaneous transplants in immune‐deficient mice. Compared to transgenic animal models that reflect the pathology of B cell malignancies, the present in vivo assessment is of limited value, since the transplanted cell lines largely recapitulate their growth behaviour in vitro. Although the characterization of XMU‐MP‐3 by Gui et al. (2019) may be enhanced in several regards, cell lines have potential, for example, for chemical genetic analysis of BTK‐dependent cell functions (Göckeritz et al., 2017) and clarification of mechanisms of acquired inhibitor resistance (Yahiaoui et al., 2017). Since inhibitors of BCR‐associated kinases also target non‐malignant bystander cells in the micro‐environment of malignant B cells (Nguyen, Niesen, & Hallek, 2019), there is plenty of opportunity for inhibitor assessment using appropriate cell line models of these cell types.