I would recommend window functions to get the maximum date:

SELECT SUM(t.total_amt) as net_total_revenue
FROM (SELECT t.*,
             MAX(t.tran_date) OVER () as max_tran_date
      FROM Transactions t
     ) t JOIN
     Customer c
     ON t.cust_id = c.customer_Id
WHERE t.tran_date >= DATEADD(day, -30, t.max_tran_date) AND 
      t.tran_date >= DATEADD(YEAR, 25, c.DOB) AND
      t.tran_date < DATEADD(YEAR, 31, c.DOB);

Some notes:

• Qualify all column names so it is clear where they come from.
• DATEDIFF() does not do what you think it does. It counts the number of Jan 1sts between two dates. DATEADD() is more accurate.
• Don't name column aliases with spaces. Use names that don't need to be escaped.

Tried to do it through temp tables

select  t.cust_id,        
sum(convert(float,t.total_amt)) as net_sales  
into #temp4  
from  Transactions t  
where t.cust_id in (select c.customer_id from Customer c where DATEDIFF(YY,DOB,getdate()) between 25 and 36)  
group by t.cust_id,t.tran_date  
having t.tran_date > DATEADD(DAY, -30, max(t.tran_date))   
order by t.tran_date desc  
GO  
select t1.cust_id, sum(net_sales) as net_sales from #temp4 t1     
group by t1.cust_id    
order by sum(net_sales) desc      
drop table #temp4