We have constructed an Escherichia coli-based platform producing (S)-reticuline, an important intermediate of benzylisoquinoline alkaloids (BIAs), using up to 14 genes. (S)-reticuline was produced from a simple carbon source such as glucose and glycerol via L-DOPA, which is synthesized by hydroxylation of L-tyrosine, one of the rate-limiting steps of the reaction. There are three kinds of enzymes catalyzing tyrosine hydroxylation: tyrosinase (TYR), tyrosine hydroxylase (TH), and 4-hydroxyphenylacetate 3-monooxygenase (HpaBC). Here, to further improve (S)-reticuline production, we chose eight from these three kinds of tyrosine hydroxylation enzymes (two TYRs, four THs, and two HpaBCs) derived from various organisms, and examined which enzyme was optimal for (S)-reticuline production in E. coli. TH from Drosophila melanogaster was the most suitable for (S)-reticuline production under the experimental conditions tested. We improved the productivity by genome integration of a gene set for L-tyrosine overproduction, introducing the regeneration pathway of BH4, a cofactor of TH, and methionine addition to enhance the S-adenosylmethionine supply. As a result, the yield of (S)-reticuline reached up to 384 mu M from glucose in laboratory-scale shake flask. Furthermore, we found three inconsistent phenomena: an inhibitory effect due to additional gene expression, conflicts among the experimental conditions, and interference of an upstream enzyme from an additional downstream enzyme. Based on these results, we discuss future perspectives and challenges of integrating multiple enzyme genes for material production using microbes.