The application of microbial and plant photosynthesis for biobased production on the one hand has a huge potential but on the other hand photosynthesis has serious limitations regarding its efficiency. Hence, studying both fundamental features of photosynthetic processes and engineering of photosystems is of paramount interest, exploring the engineering of photosystems is the overarching aim of this thesis. As described in Chapter 1, natural photosystems may be modified or transplanted to allow for more efficient conversion of solar light energy into biochemical energy. Hereto ambitious proposals to engineer photosystems have been made, and to realize those endeavors the disciplines of synthetic and systems biology are required. To explore how to apply and improve those disciplines hereto, the work described in this thesis has focused on the transplantation of simple photosystems (proton-pumping rhodopsins; PPRs) into the cell membrane of the heterotrophic model bacterium Escherichia coli. Both in silico analyses, including metabolic and thermodynamic modeling (Chapter 3) and a series of experimental studies on transplanting PPR photosystems (Chapters 4,6 and 7) were performed, which identified several challenges, limitations and most importantly opportunities. This thesis also describes the application of novel tools to substantially improve the functional production of PPRs and a variety of other membrane proteins in E. coli. Chapter 2 provides more details on previously reported examples of heterologous expression of PPRs in several hosts, and on the physiological impact of these transplanted photosystems. Based on this evaluation, some suggestions are made to improve and further exploit the transplantation of these photosystems. In Chapter 3 a systematic, integrated in silico analysis is made of anaerobic, photo-electro-autotrophic synthetic metabolism in E. coli, consisting of (i) a PPR photosystem for ATP regeneration, (ii) an electron uptake pathway, and (iii) a natural or synthetic carbon fixation pathway. Constraint-based metabolic modelling of E. coli central metabolism is used, in combination with kinetic and thermodynamic pathway analyses. The photo-electro-autotrophic designs are predicted to have a limited potential for anaerobic, autotrophic growth of E. coli, given the relatively low ATP regenerating capacity of the PPR photosystems, and the relatively high ATP consumption due to maintenance. In general these analyses illustrate the potential of in silico analyses to identify potential bottlenecks and solutions in complex designs for autotrophic and photosynthetic metabolism, as a basis for subsequent experimental implementation. To tackle a main bottleneck of PPR systems: their functional membrane-embedded production level, the heterologous production in E. coli of the proton-pumping rhodopsins from Gloeobacter violaceus (GR) and from Thermus thermophilus JL18 (TR) is quantified and experimentally optimized in Chapter 4. High constitutive production of both rhodopsin proteins is achieved by fine-tuning transcription and translation. Besides the canonical retinal pigment, the GR system has the ability to bind a light-harvesting antennae pigment, echinenone. After optimization of the heterologous pigment biosynthesis pathways for either retinal or echinenone production, appropriate quantities of retinal or echinenone for PPR reconstitution were detected in E. coli. Association of echinenone with GR broadens its absorption spectrum in E. coli, broadening the potential for light-harvesting also to blue light. Optimization of the branched pathway for simultaneous biosynthesis of both retinal and echinenone has been attempted by using a smart library of variable Ribosome Binding Sites (RBSs) with varying strengths (RedLibs). In general, the here described approaches towards improved functional production of rhodopsin photosystems in E. coli and their pigments may prove more widely applicable for heterologous production of more complex photosystems and other systems. In Chapter 5 an up-to-date overview is provided on how codon usage can influence functional protein production. The fact that all known organisms have an incomplete set of tRNAs, indicates that biased codon usage could act as a general mechanism that allows for fine-tuning the translation speed. Although translation initiation is the key control step in protein production, it is broadly accepted that codon bias, especially in regions further downstream of the start codon, can contribute to the translation efficiency by tuning the translation elongation rate. Modulation of the translation speed depends on a combination of factors, including the secondary structure of the transcript (more or less RNA hairpins), the codon usage landscape (frequent and more rare codons) and for bacteria also RBS-like sequences at which ribosomes can pause. The complex combination of interdependent factors related to codon usage that can influence translation initiation and elongation. This complexity makes that the design of synthetic genes for heterologous expression is still in its infancy, and despite the availability of some codon usage algorithms, it is often as well a matter of trial and error. In Chapter 6 the effect of different codon usage algorithms (optimization and harmonization) has been experimentally tested for heterologous production of membrane proteins. Apart from the codon usage algorithms also the combined effect of transcriptional fine-tuning in E. coli LEMO21(DE3) was assessed. The overproduction of 6 different membrane-embedded proteins, including 4 PPR variants (from bacteria, archaea and eukaryotes), was tested. For production of tested PPR variants, the different codon usage algorithms hardly influenced production, while transcriptional tuning had a large impact on production levels. Interestingly, for the other two tested non-PPR membrane proteins, some codon usage variants significantly improved production on top of transcriptional tuning. For both these proteins the codon-optimization algorithm reduced functional production below that of the wild-type codon variant, while the harmonization algorithm gave significantly higher production, equal or even higher than for the wild-type variant. In Chapter 7 it is demonstrated that a translational-tuning system can be used to successfully optimize the expression of several membrane proteins, based on initial findings presented in Chapter 4. The employed, recently developed Bicistronic Design (BCD) system is based on translational coupling of a gene encoding a short leader peptide and the gene of interest that is under control of a variable ribosome binding site. A standardized library of 22 RBSs allows for precise, gene context-independent, fine-tuning of expression of this second gene, here encoding a membrane protein. For all four membrane proteins tested in this study the BCD approach resulted in 3 to 7-fold higher protein levels than those obtained by two other recently developed methods for optimizing membrane protein production. The presented approach allows for inducer-free, constitutive, high-level production of membrane proteins in E. coli, which can be widely applicable for both membrane protein purification studies as well as for synthetic biology projects involving membrane proteins. In Chapter 8 a broad review and perspectives are provided on the potential of microbial autotrophs for the production of value-added compounds from CO2. Both photoautotrophic and chemolithoautotrophic production platforms are discussed, and recent progress in improving their efficiency and production potential is highlighted. Transplantation efforts for photosystems, but also for CO2 fixation pathways and electron uptake systems are discussed. An overview is provided on novel in silico and experimental approaches to engineer components related to autotrophy in heterotrophic and autotrophic model hosts, including approaches applied in this thesis. Future avenues are discussed for realizing more efficient autotrophic production platforms. Finally, in Chapter 9 and 10 the work in this thesis is summarized and a general discussion is provided on future avenues for engineering of PPR photosystems, photosystems in general and on the optimization of membrane protein production.