Black's best recorded variations for this line (plan in variation A) gives good ideas):

6.Qb3 Nc6 

A) 7.Qxb6 axb6 8.Na3 Na7!? 9.Nc2 (9.Be2 cxd4 10.cxd4 Bxa3 11.bxa3 Bb5=) Bd7 10.Bxb5+ Nxb5 11.Be3 Nh6= or 11...Na7 12.oo Rc8 13.a4 Nc6

Black can be confident in these positions with a light squares control

B) 7.Na3 Na5 (need serious attention, because you can transpose to this position from A))  8.Qxb6 axb6 9.Be2 (9.Bd3!?...and here I think Black can continue his plan withNc8-Na7 and Bb5 ) Aseev-Rustemov,St Petersburg 1997 (Jovicic)  9... Nac6!?  (or 9...Nc8 with the idea Na7 and Bb5) 10.oo f6 (Rustemov and Jovicic)

In these lines Black continue fight on queenside (basic pattern when White has no more attack in center or kingside)

C) 7.Be3 Lupu-Legky,Besancon 1998 7...cxd4 8.cxd4 (8.Bxd4 Nxd4 =+) Qxb3 (now White has permanent weakness with a double pawn on b2-b3 and Black has strong control of b4 square) 9.axb3 Nge7 10.Nc3 Nf5 11.Be2 Be7 12.Nb5 oo -/+

D) 7.Bd3 Qxb3 9.axb3 cxd4 =+ same positional ideas than C)

Per the recommendation of Neil McDonald in "How to Play Against 1. e4", I've been trying 1. e4 e6 2. d4 d5 3. e5 c5 4. c3 Qb6 5. Nf3 Bd7, with Bb5 soon to trade black's bad bishop for white's good one. 

In a couple of recent games, I've had opponents respond with an early Qb3, offering a queen trade before I get a chance to play Bb5. The first time, I worried about what white might be able to do with the semi-open a file after Qxb3 axb3, so I didn't initiate the trade. My opponent traded queens on b6, giving me a semi-open a file, but it didn't help my rook any. 

In my most recent game in this variation, I traded queens on b3, then traded c pawns on d4, leaving him with doubled, isolated b pawns, which seems good for me. But I got in a little trouble, because I couldn't use a6 to prevent Nb5, because my a pawn was pinned to my unguarded rook on a8.

I still think trading on b3 was probably the right thing to do. Anyone have thoughts on preventing that Nb5 maneuver in this variation?