Function of Rac

1.3.3.1 Rac and the actin cytoskeleton

M icroinjection of constitutively activated V12 or L61 Rac into confluent, serum -starved Swiss 3T3 cells results in actin polym erisation at the plasm a m em brane leading to lam ellipodia

and membrane ruffles (Ridley et aL, 1992). Some growth factors,

such as PDGF, EGF, thrombin and insulin, have a similar effect which can be blocked by microinjection of N17 Rac, a dom inant negative form locked in a GDP-bound state (Ridley et aL, 1992). This has led to the proposal that Rac controls a growth factor- m ediated signal tran sd u ctio n p ath w ay lead in g to lam e llip o d ia formation. Lam ellipodia are characteristic o f m otile cells and are thought to be required for cell movement, suggesting that Rac is an im p o rta n t p lay e r in cell m o v em en t (M ach esk y and Hall, 1997). R u ffles are d eriv ed from d e ta ch e d la m e lla e and are believed to play a part in pinocytosis.

The ruffling response to growth factor stim ulation or to Rac injection is observed after 5 m inutes. It was noted that at later tim e points (20-30 m inutes), the cells start to develop stress fibres which can be blocked with C3. This suggested that.

as well as stimulating m em brane ruffling, Rac also leads to the activation of Rho, albeit more weakly than direct activation by L PA . It has b e en r e p o r te d th at R a c - m e d ia te d le u k o tr ie n e

production accounts for the activation o f Rho (Peppelenbosch e t

a l , 1995).

M icroinjection o f V12 Ras has been long reco g n ised to cause m em brane ruffling (Bar-Sagi and Feram isco, 1985). It is now clear that this is also due to cro ss-talk b etw een small GTPases and that Ras activates Rac (Ridley et a l , 1992). In Swiss 3T3 cells, Ras is not required for Rac activation by PDGF, EGF or insulin (Ridley et a l , 1992).

Activation of Rac by PD GF in Swiss 3T3 cells is dependent on P I3K activity, since PD G F re c e p to r m u tan ts th at can n o t interact with PI3 kinase can no longer induce ruffling. Also, the PI3 kinase inhibitor w o rtm an n in blocks m em b ran e rufflin g in response to PD GF or insulin (W ennstrom et a l , 1994, Kotani e t a l , 1994). H ow ever, w ortm annin has no effect on m em brane ruffling induced by microinjection of activated Rac, showing that PI3 kinase activity is required upstream o f Rac (Nobes et a l ,

1995). Reports that PDGF stimulates the formation of Rac-GTP in

a PI3 kinase-dependent m anner confirm this m odel (Hawkins e t

a l , 1995). Nobes et al reported that wortm annin did not inhibit R a s - in d u c e d m e m b ra n e r u f f lin g in S w iss 3T3 c e lls , b u t R o d r ig u e z - V ic ia n a et al h av e gone on to show c o m p le te in h ib itio n o f Ras ru fflin g w ith a d o m in a n t n e g a tiv e P I3 K construct in Porcine Aortic Endothelial (PAE) cells (Nobes et a l ,

1995, R odriguez-V iciana et a l , 1997). This raises the possibility th at Ras can activate R ac th rough a w o r tm a n n in -in s e n s itiv e PI3K. An interaction between the p85 subunit of PI3 kinase and Rac, requiring the rhoGA P dom ain of p85, has been described (Zheng et a l , 1994). This interaction is G TP-dependent and may activate the PI3 kinase activity.

Rac, like Rho, also induces the form ation o f integrin-based focal adhesion complexes at the plasm a m em brane (N obes and Hall, 1995). These focal complexes are similar in com position to classical focal adhesions (e.g. they also contain vinculin, paxillin and p p l2 5 F A K ), but they have a distinct m orphology; they are s m a lle r and d i s tr ib u te d a lo n g th e e n tir e le n g th o f th e

lam ellipodia edge. They are form ed in dependently o f Rho and their function is unknown. H ow ever, like Rho, the actin and adhesion com plex pathw ays appear to be separable, since Rac- m e d ia te d a ctin p o ly m e ris a tio n still o ccu rs u n d e r c o n d itio n s where focal complexes cannot form (such as in cells plated on po ly -L -ly sin e) (H otchin and Hall, 1995, M a ch e sk y and Hall,

1 9 9 7 ).

In a d d itio n to th e r e p o r ts d e s c r i b i n g R a c - i n d u c e d lam ellipodia form ation in Swiss 3T3 cells, sim ilar effects have been found in other cell types, for exam ple the epiderm al cell line KB (Nishiyama et al., 1994), in M D CK epithelial cells (Ridley et al., 1995b), in m ast cells (N o rm an et al., 1994) and in m acrophages (Allen et al., 1997). Rac, like Rho, is necessary for establishm ent of cadherin-based cell-cell contacts (B raga et al.,

1997) and has also been reported to be involved in endocytosis (Lam aze et al., 1996) and regulated exocytosis (O'Sullivan et al.,

1 9 9 6 ).

1.3.3.2 Rac, stress kinases and control o f trancription

P erh ap s the m ost in te re s tin g d e v e lo p m e n t in the Rho GTPase field in the past couple of years has been the discovery th at th ese p ro te in s c o n tro l gene t r a n s c r ip tio n v ia se v e ra l d is tin c t p a th w a y s (V an A e lst and D 'S o u z a - S c h o r e y , 1997). A lthough it is not clear at this stage w hat genes are directly controlled by Rho, Rac and Cdc42, it seems that, in response to e x tra c e llu la r stim u li, they c o -o rd in a te ly r e g u la te lo n g term c h a n g e s i n v o lv in g g en e e x p re s s io n w ith th e s h o r t term cytoskeletal rearrangem ents described above. All o f the studies on transcription control by m am m alian Rho GTPases have so far involved transfection o f tissue culture cells, though recent w ork in D r o s o p h i l a has shown that Rho GTPases do indeed control gene transcription in vivo (see 1.3.6 .4 for a detailed discussion). JNK/SAPK and p38/RK

M A P K in a se p a th w ay s are c o n s e rv e d m o d u la r k in a s e c a s c a d e s th a t m e d ia te sig n a l t r a n s d u c t io n e v e n ts in all eu k ary o tes (H ersk o w itz, 1995, M a rsh a ll, 1994, K y riak is and A vruch, 1996). The p rototypical M A P K inase path w ay is the R a s - c o n tr o lle d E R K l / 2 p a th w a y o f m a m m a ls. In a d d itio n

mammals have at least two other MAP Kinase pathways, the JNK ( c -ju n N - te r m in a l K in a s e ) /S A P K (S tre s s a c tiv a te d p r o te in kinase) and p38/RK (Reactivating Kinase) pathw ays, also known as stress response kinases (review ed by K yriakis and Avruch, 1996). G row th facto rs th at stim u late the E R K p a th w a y are m ostly p oor activators o f the stress response kinases, though EOF has been reported to give a 15-fold activation of JN K in HeLa cells (Minden et al., 1995). In contrast, JN K and p38 are both comm only activated in response to cellular stresses such as ultraviolet irradiation or heat and inflam m atory cytokines such as T N F a and in te rle u k in -1 (IL-1). These kinase pathw ays are, therefore, th ought to play a role in inflam m ation. Figure 1.6 sum m arises the kinases involved that have been cloned so far and their know n transcription factor targets. (T reism an, 1995, Treisman, 1996, Hill and Treism an, 1995). It can be seen that, for the JN K and p38 pathways, a large num ber o f isoform s of each kinase have been found and the overall p ictu re is m ade more com plex by functional redundancy and crosstalk betw een the d ifferen t pathw ays.

W hile the R as-controlled ERK pathw ay is well understood, until a few years ago, events linking stress and cytokines to the JN K and p38 pathways were unclear. Activated versions o f Rac and Cdc42 have now been reported to stim ulate both the p38

and JNK cascades in Cos, NIK 3T3 and HeLa cells (Coso et al.,

1995, M inden et al., 1995, B agrodia et al., 1995, Olson et al., 1995, Zhang et al., 1995). In 293T cells, Cdc42 and Rho, but not

Rac, appear to stimulate JN K and p38 (Teramoto et al., 1996a).

Thus it would appear that there are cell type differences in Rho GTPase regulation of stress kinases. All these experim ents have been done with overexpressed GTPases and kinases.

GTPase