Applications of the reverse pharmacology strategy

From the discovery that the ligand for the orphan receptor G-21 was serotonin (Fargin etal. 1988), which represented the first example of'ligand pairing', the methodology has not really evolved appreciably per se. The greatest advances have been in the 'technology', for example, sensitivity of the assays to increase signal-to-noise, and the miniaturization of assay formats to allow high-throughput. Nowadays, it is very much the standard to be screening in 384 well format, but in general, we continue to use changes in cAMP levels or mobilization of intracellular calcium as endpoints. Other, more generic assays have been applied, making use of natural phenomena such as the measurement of colour changes using frog melanophores

(Birgul et al. 1999; Lynch et al. 1999; Lenz et al. 2001) or changes in extracellular pH using the microphysiometer (Habata et al. 1999; Hinuma et al. 2000).

The advantage of the melanophore technology is that the pigment bearing cells contain all the major G proteins (Gq, Gs, and Gj) allowing a functional response to be elicited without detailed knowledge of the receptor's coupling mechanism. In melanophores, Gs or Gq activation results in pigment dispersion, while Gi mediated signalling leads to pigment aggregation. The technology is ideally suited for use in multi-well screening and pharmacological profiling of lead compounds. However, the technology is not considered to be high-throughput by today's standards. The attractiveness of this technology for many of the major pharma groups is reduced due to the difficulty in obtaining melanophores, the current need for high-throughput assay formats, and the ease-of-use and generic nature of today's mammalian expression systems.

The microphysiometer, which measures extracellular media acidification in response to perturbations of energy consumption in mammalian cells, provides a low-throughput capability. Since activation of any signal-transduction pathway results in energy usage, this approach also obviates the need to define intracellular signalling mechanistically. This assay has been used to identify the natural ligand for the orphan GPCR APJ, termed apelin (Habata et al. 1999), and the receptor for mammalian RF-amide peptides (Hinuma et al. 2000). However, in the absence of improvements for a higher throughput capacity assay, broader application of the above technologies will be limited.

The current orphan strategy makes use of promiscuous and chimeric G proteins which allow orphan receptors to be functionally coupled through a common pathway, resulting in calcium mobilization via phospholipase C (PLC) in standard mammalian cell-lines. While unlikely to work for all GPCRs, it has considerably broadened the range of receptors that will give measurable calcium mobilization responses. The approach makes use of the naturally occurring promiscuous G proteins (Ga15 and Ga16) (Offermanns and Simon 1995), and chimeric Ga subunits (Gqi5), in which the C terminal five amino acids of the Gq subunit are replaced by corresponding amino acids from the adenylyl-cyclase linked Gi subunit (Milligan and Rees 1999). The Gq family subunit, Gq11, was used in the discovery of the ghrelin receptor (Howard etal. 2000) and more recently, several groups identified the melanin-concentrating hormone 1 and 2 (Chambers et al. 1999; Sailer et al. 2001), and the NPFF1 (Elshourbagy et al. 2000) receptors using Ca2+-based assays in which chimeric G proteins were used.

To date, the most successful strategy to identify ligands for orphan receptors has been accomplished using assays that report changes in intracellular Ca2+ as a result of PLC activation. Changes in calcium mobilization can be easily detected using standard fluorescent based methods, via a high-throughput imaging system such as FLIPRâ„¢ (Fluorescent Imaging Plate Reader, Molecular Devices). Several important biological mediators defined by this means are summarized in Table 10.1. The advantages of these assays are their high-throughput capability (automated and 96 or 384-well plate capacity), robustness (high signal-to-noise ratio arising from large intracellular amplification) and flexibility (mode of detection and varied instrumentation). This technology has proved successful in our labs, with a number of ligand pairs identified. For example, we and others, have found the endogenous ligand for FM-4 to be Neuromedin U (Howard et al. 2000) and for HG57 to be LTC4 (Heise et al. 2000). These being the NUM2R and CysLT2 receptors, respectively.

As already stated, the use of promiscuous G proteins has allowed the effective coupling of Gq and Gi receptors to calcium mobilization, but the ability to couple those receptors that stimulate adenylyl-cyclase via the G protein Gs to calcium mobilization has proven much

Table 10.1 Identification of ligands for orphan GPCRs. The ligand pairings are grouped on the basis of the mechanism by which they were discovered. The citation is only illustrative and not comprehensive as in a number of cases several groups have identified the same ligand pair. An attempt to cite the first publication or back-to-back publications for each of the receptors has been made.

Table 10.1 Identification of ligands for orphan GPCRs. The ligand pairings are grouped on the basis of the mechanism by which they were discovered. The citation is only illustrative and not comprehensive as in a number of cases several groups have identified the same ligand pair. An attempt to cite the first publication or back-to-back publications for each of the receptors has been made.

Measured

Orphan GPCR

Ligand

Putative function

References

activity

paired

Ca2+

HFGAN72

Orexin

Feeding/sleep-wakefulness

Sakurai etal. 1998

GHS-R

Ghrelin

GH secretion/Adiposity

Kojima etal. 1999

GPR-9-6

TECK

Gastric inflammation

Zaballos etal. 1999

GPR24

Melanin-concentrating

Feeding

Chambers etal. 1999;

hormone

Saito etal. 1999

HG55

Leukotriene D4

Bronchial constriction

Lynch et al. 1999

GPR38

Motilin

Gastric motility

Feighner etal. 1999

GPR14

Urotensin II

Vasoconstrition

Ames et al. 1999

OGR-1

Spingosylphosphorylcholine

Cell proliferation

Xu etal. 2000

HG57

Leukotriene C4

Bronchial constriction

Heise etal. 2000

444-J19

Leukotriene B4

Bronchial constriction

Wang etal. 2000

GPR2

Eskine

Leukocyte traffiking

Jarmin etal. 2000

FM-3

Neuromedin U

Feeding

Fujii et al. 2000

FM-4

Neuromedin U

Feeding

Howard etal. 2000

HLWAR77

NPFF

Pain modulation

Elshourbagyetal. 2000

SP1999

ADP

Thrombosis

Zhang et al. 2001

SLT

Melanin-concentrating

Obesity

Sailer etal. 2001

hormone

CRTH2

Prostaglandin D2

Induces TH2 chemotaxis

Hirai etal. 2001

GPR54

KiSS-1

Trophoblast Invasion

Ohtaki et al. 2001

G2A

Lysophosphatidylcholine

Kabarowski etal. 2001

GPR4

Spingosylphosphorylcholine

Atherosclerosis/

Zhu etal. 2001

inflammation

cAMP

ORL-1

Nociceptin

Anxiety/memory

Reinscheid etal. 1995

EDG-1

Spingosine-1-phosphate

Cell differentiation/

Lee etal. 1998

growth

GPCR97

Histamine

CNS-obesity, psychiatry

Lovenberg et al. 1999

H4

Histamine

Inflammation

Oda etal. 2000

TDAG8

Psychosine

Cytokinesis

Im etal. 2001

GPR86

ADP

Hematopoiesis/

Communi et al. 2001

inflammation

TA1

Tyramine

Depression/electrolyte

Borowskyetal. 2001

homeostatis

Energy

APJ

Apelin

Immune response

Habata etal. 1999

utilization

OT7TO22

RF-amides 1-3

Pain

Hinuma et al. 2000

GIRK channels

DAR-1

Allostatin

?

Birgul etal. 1999

DAR-2

Allostatin

?

Lenz etal. 2000

Arach. acid

GPR10

Prolactin releasing peptide

Prolactin secretion

Hinuma etal. 1998

Yeast

EDG-2

Lysophosphatidic acid

Cell differentiation/

Erickson etal. 1998

pheromone

growth

KIAA0001

UDP-glucose

?

Chambers etal. 2000

more difficult (Milligan and Rees 1999). However, the availability of assays that measure the cytoplasmic level of cAMP via a reporter gene construct, for example, making use of the cAMP response elements (CRE) upstream of either ^-lactamase or Luciferase reporter genes, has provided a sensitive and reasonably straightforward assay to allow high-throughput screening. We have made extensive use of this assay in parallel to the FLIPRâ„¢ screen as our primary screening cascade and it has delivered a number of successes. For example, we and others have identified an orphan GPCR that is highly expressed in the kidney and is activated by tyramine and P-phenylethylamine; this being the first member of the Trace Amine Receptor family (Borowsky et al. 2001). Other alternatives to the reporter-based methods for measuring cAMP levels include radioimmunoasay and scintillation proximity assay (Amersham Pharmacia Biotech). These assays have been used to identify the ligands for the orphan receptor GPR97 as the Histamine H3 receptor (Lovenberg et al. 1999) and GPR86 as the ADP receptor (Communi etal. 2001) (Table 10.1).

Functional expression of GPCRs in yeast offers an alternative high-throughput option for screening. The pheromone pathway has been engineered to permit functional coupling of human GPCRs to growth on a selective media or reporter gene expression (Price et al. 1995). In comparison to mammalian cells, yeast cells are robust, cheap and straightforward to handle. This methodology has had limited success (Erickson et al. 1998; Chambers et al. 2000) compared to the other systems described above. This is probably attributable either to the difficulty in demonstrating functional expression of the GPCR in yeast or to ligand binding to components of the yeast cell wall.

In industry, the ability to screen in a high-throughput manner is key as it is essential to screen an orphan receptor against a wide range of putative ligands. We have amassed a collection of over 2000 naturally occurring putative GPCR ligands (bioactive peptides, small molecules, lipids, and carbohydrates) which these constitute the first phase in the screening process. This 'ligands file' contains all known or suspected GPCR ligands, together with a large number of molecules for which the receptor is unknown. It contains a number of bioactive peptides from lower species and novel peptides derived through in-silico analysis of the human genome. Screening against this ligand file has provided us with a number of successes. If no hits are identified we initiate a second round of screening for novel ligands using biological extracts of tissues, fluids, and cell supernatants. In addition we may also screen against a file of small organic molecules, selected computationally based on known GPCR pharmacophores, in order to identify surrogate ligands. These can be used as tool compounds to explore receptor function in more physiological systems. Once a ligand has been identified a high-throughput screen is initiated to identify antagonists of the receptor.

0 0

Post a comment