to amplify the PERV-B(33) backbone in two steps leaving four homologous regions to be ligated during Gibson assembly. Amplification of fragments was successful and clear bands were observed as seen in figure 3.20. However efforts have so far been unsuccessful to assemble the various fragments after gel extraction via Gibson assembly. Possible reasons could be in the degradation of DNA over time, as these fragments were not used immediately. Another reason may be a too short extension period for the long (6 kb) amplicon. This may result in variable-length fragments that could be observed as slightly thicker bands on a 1 % agarose gel (Figure 3.20). At last, genetic material recovered may have been contaminated with salts or other organic compounds due to the elution process from the agarose gel. To circumvent this possibility in the future and to improve cloning efforts, a DpnI digest could be used to digest methylated template DNA only leaving intact amplicons. This is possible, as no extra bands were obtained. The resulting solution could then be cleaned via PCR-clean up kit (Invisorb ® DNA CleanUp A.5) to minimize the effect of residual ethanol and salts for downstream applications. It seems that a viable system was found to bypass the problem, with all tools in place to create a replication competent clone, provided the constructs given allow for proper assembly of the virus. Similarly the production of protein expression vectors, wild type and mutated, has revealed additional problems, such as long length amplicons and good efficiency in terms of protein secretion. The employment of a well functioning site-directed mutagenesis technique as well as the proper incorporation of the mutated fragments into the expression vector is crucial for success. Additionally, Env contains a transmembrane domain, which anchors the protein within the membrane of a cell, making it impossible to isolate sufficient amounts of Env from the supernatant of cells in cell culture(Figure 2.3). In order to prevent this from happening the env gene needed to be truncated in a way that it’s transmembrane domain was cleaved out and it could be readily secreted from the cell. Otherwise only harvesting of cell-lysate, which contains remnants of cellular debris, would have been possible. Such debris (e.g. fibrinogen and heparan sulfate) may be a reason for false readings when conducting IL-10 assays as it can also trigger TLR4 [54, 55].
While most pathogens can be eliminated by specified pathogen-free breeding, this derivation technology is not expected to affect the potential transmission ofporcineendogenousretroviruses (PERV) as they are an integral part of all pig genomes. Generally, trans-species transmissions occur in nature and result either in immunodeficiencies (e.g., human immunodeficiency viruses: HIV-1 and HIV-2) and/or tumors (e.g., feline leukemia virus, FeLV; Koala retrovirus, KoRV) . Although PERVs are able to infect several human cell types in vitro [8–12], the potential to infect humans in vivo and possibility to cause diseases remains to be established. Exploratory clinical xenotransplantation trials including ex vivo perfusion allowing short-term contact of human recipients to living porcine cells or tissues have not resulted in transmission of PERV [13–21]. Similarly, investigations of non-human primates that had received porcine cells or organs showed no
Human pathogens that can be transmitted by allotransplantation are in most cases well characterised, for example HIV-1, HEV, rabies virus, and sensitive detection methods are generally available. However, in the case ofxenotransplantation, the potentially zoonotic microorgan- isms remain partially unknown and sensitive detection methods are under development. There are indeed a lot of veterinary diagnostic lab- oratories with assays for infectious disease monitoring in pig herds, however their assays are usually monitoring pig diseases with high im- pact on meat production. Based on early ﬁndings showing that some de- tection methods are not sensitive enough  , new and improved methods for potentially zoonotic microorganisms are under develop- ment. There is evidence that some porcine viruses are able to infect human cells in vitro or humans in vivo, e.g., HEV. Whereas the microor- ganisms transmitted after allotransplantation are adapted to humans, many porcine microorganisms fail to infect humans due to the innate and adaptive immune response, due to the absence of suitable receptors on human cells or due to human intracellular restriction factors. Restriction factors are cellular proteins that inhibit viral replication and represent a ﬁrst line of defense against viral pathogens  . They show an enormous structural and functional diversity and target nearly every step if the viral replication cycle. Examples of such restriction fac- tors are APOBEC3 (apolipoprotein B mRNA editing enzyme catalytic polypeptide 3), which induces hypermutation by deamination, or tetherin, which prevents virus release by tethering budding progeny vi- rions to the plasma membrane of the infected cell  . Bacteria, fungi, parasites and viruses are potentially zoonotic porcine microorganisms. Whereas bacteria and fungi can be treated and eliminated by antibiotics and antifungals, treating viral infections with antivirals is more compli- cated and it seems that therefore viruses represent a greater risk. Among the potentially zoonotic viruses, porcineendogenous retrovi- ruses (PERVs), porcine cytomegalovirus (PCMV), HEV genotype 3, por- cine lymphotropic herpesviruses (PLHV), andporcine circoviruses (PCV) are the best characterised ( Table 1 ). Whereas PCMV, HEV, PLHV, PCV as well as bacteria and fungi may be eliminated from pigs
Die Xenotransplantation mit porzinen Zellen, Gewebe oder Organe könnte die immer größer werdende Lücke zwischen Nachfrage und Angebot der menschlichen Spenderorgane reduzieren. Dies kann jedoch durch die Anwesenheit von porzinen endogenen Retroviren (PERVs) erschwert werden. PERVs gehören zu den γ-Retroviren und sind im Genom aller Schweine integriert und können menschliche Zellen in vitro infizieren. Drei Subtypen des PERVs wurden identifiziert, PERV-A und PERV-B sind ubiquitär und besitzen die Fähigkeit humane Zellen in vitro zu infizieren. Daher stellen sie ein unmittelbar Risiko für die Xenotransplantation dar. PERV-C infizieren nur Schweinezellen und sind nicht in allen Schweinestämmen vorhanden. Bis dato, wurde kein Beweis über eine in-vivo-Übertragung von PERVs auf Menschen oder nichthumanen Primaten nach Xenotransplantation von Schweinmaterial erbracht. Dennoch, viele der γ-Retroviren sind pathogen und können Tumore und Immundefizienzen verursachen und dies kann auch für PERVs nicht ausgeschlossen werden. Mit der Aufzucht der Schweinen unter DPF (designated pathogen free) Bedingungen kann es gelingen, die meisten Krankheitserreger und Viren zu eliminieren, was im Fall von PERVs nicht möglich ist, da PERV Proviren bis 100-mal im porzinen Genom integriert sind. Daher müssen andere Strategien verwendet werden, um eine möglich Übertragung auf den Menschen zu verhindern.
While most of these pathogens are sufficiently controlled by housing animals within barrier facilities and thus protecting them from exogenous infections, porcineendogenousretroviruses (PERVs) are a more precarious issue as they can be transmitted via the germline. Consequently, PERVs have caused considerable concern for clinical xenotransplantation. But to date there has been no report of PERVs being actually transmitted to the recipient of a xenograft or having caused detectable adverse reactions within a transplant 10 . Recent successes in eliminating PERVs from the genome of pig lines have further diminished this potential hazard 11,12 . Other pathogens, however, provoke lifelong, latent, transmittable infections, and are prevalent in most, if not all, pig populations and are thus difficult to control. These include the porcine cytomegalovirus (PCMV), a β-herpesvirus related to the human cytomegaloviruses that causes systemic disease and potentially leads to graft failure in human allotransplantation 13 . In xenotransplantation, the PCMV has also been associated with transplant injury. This has been largely attributed to virus activation within the graft following transplantation but initially was not thought to cause invasive disease in the recipient 14,15 . It is still unclear whether PCMV can infect human cells, with one in vitro study suggesting the possibility 16 and a different study presenting evidence for the opposite 17 . There are indications that while in vitro PCMV appears susceptible to standard antiviral medication comparable to that employed in allotransplantation 18 , in vivo data from pig-to-baboon xenotransplantation suggest that the commonly used ganciclovir has no therapeutic efficacy against PCMV at standard doses 19 . Agents that do prevent or treat PCMV infection effectively, such as foscarnet or cidofovir, carry significant toxic potential for the transplant recipient and are thus of limited usefulness. Consequently, there is consensus that potential donor animals for xenotransplantation should be free of PCMV.
Xenotransplantation has been proposed as a solution to the shortage of suitable human donors for transplantation and pigs are currently favoured as donor animals. However, xe- notransplantation may be associated with the transmission of zoonotic microorganisms. Whereas most porcine microorganisms representing a risk for the human recipient may be eliminated by designated pathogen free breeding, multiple copies ofporcineendogenousretroviruses (PERVs) are integrated in the genome of all pigs and cannot be eliminated this way. PERVs are released as infectious particles and infect human cells. The zinc finger nu- clease (ZFN) technology allows knocking out specifically cellular genes, however it was not yet used to eliminate multiple integrated proviral sequences with a strong conservation in the target sequence. To reduce the risk of horizontal PERV transmission and to knock out as many as possible proviruses, for the first time the powerful tool of the ZFN technology was used. ZFN were designed to bind specifically to sequences conserved in all known rep- lication-competent proviruses. Expression and transport of the ZFN into the nucleus was shown by Western blot analysis, co-localisation analysis, PLA and FRET. Survival of trans- fected cells was analysed using fluorescent ZFN and cell counting. After transfection a strong expression of the ZFN proteins and a co-localisation of the expressed ZFN proteins were shown. However, expression of the ZFN was found to be extremely toxic for the trans- fected cells. The induced cytotoxicity was likely due to the specific cutting of the high copy number of the PERV proviruses, which is also commonly observed when ZFN with low specificity cleave numerous off-target sites in a genome. This is the first attempt to knock out multiple, nearly identical, genes in a cellular genome using ZFN. The attempt failed, and other strategies should be used to prevent PERV transmission.
Whereas the viruses discussed above may be eliminated by selection, treatment, vaccination, Caesarean delivery, early weaning and embryo transfer, this is impossible in the case ofporcineendogenousretroviruses (PERVs). These viruses are the result of infections with retrovi- ruses in the distant past. Retroviruses possess enzymes able to transcribe the viral RNA genome into DNA cop- ies and to integrate them into the cellular genome to form proviruses . If oocytes or sperm cells happen to become infected, these proviruses will be present in all cells of the developing progeny and will be inherited in the same way as all other genes. Endogenous retrovi- ruses have been found in all mammals including humans . However, in contrast to the human endogenous ret- roviruses which are mostly defective and do not produce infectious particles, many retroviruses in other species remain active. Gamma- and betaretroviruses have been found integrated into the genome of pigs [58, 59] and sequencing of the entire pig genome (Sus scrofa) re- vealed 212 PERV insertions in the genome . The gammaretroviruses include PERV-A and PERV-B, which are integrated into the genome of all pigs, and PERV-C, found in many (but not all) pigs. PERVs are produced as infectious virus particles and may infect human cells [61–63]. In addition, recombinants between PERV-A
The identification ofporcine viruses so far unrecognized is required to minimize virus-related risks associated with xenotransplantation. We used a pan-herpes consensus PCR assay to search for unrecognized porcine species of the Herpesviridae. The assay targets conserved regions of the herpesvirus DNA polymerase (DPOL) gene, using primers that were modified to diminish the assay´s recognition capacity for the highly prevalent porcine lymphotropic herpesviruses 1, 2 and 3 (PLHV-1, -2, -3), without substantially lowering the universal detection capacity of the assay. Analysis of 495 porcine blood and tissue samples from 294 animals, including 35 samples from 20 immunosuppressed pigs, resulted in the amplification of 128 herpesviral DPOL sequences. Sequence analysis attributed 127 of the amplimers to the known porcine herpesviruses (PLHV-1, -2, -3; porcine cytomegalovirus; pseudorabiesvirus). In none of the pig samples analysed here, evidence was obtained for the presence of additional novel porcine herpesvirus species. Therefore we conclude that pigs bred for the purpose ofxenotransplantation pose a negligible risk of transmitting presently unrecognized herpesviruses to organ recipients.
Retroviruses are tumorigenic and may cause immunodeficiencies in infected hosts. The risk of PERV infection has to be carefully assessed specifically in view of the fact that the human immunodeficiency viruses HIV-1 and HIV-2, causing AIDS in humans, are the result of multiple trans-species transmissions ofretroviruses from non-human primates [17,18]. However, in several clinical applications ofporcine islet cells or ex vivo perfusions of pig liver cells such as bridging during acute liver failure, no virus transmission could be detected [19–22]. Likewise, in a cohort of 160 patients treated with various porcine tissues, no PERV transmission was observed [23,24]. Moreover, PERV transmission was not found after pig to non-human primate transplantation and PERV infection experiments [25–27]. However, the risk of PERV transmission associated with tumors or immunodeficiencies is still prevalent . Removal of PERV sequences by knock-out technology would be the safest strategy, but may prove difficult due to the presence of numerous replication competent and defective proviruses capable of recombination and complementing each other.
For both workers andsafety experts, work safety perceptions are not only subjective, but also shaped by their own beliefs, knowledge, and experience.
In the case of workers, safety knowledge is primarily driven by task-performance experience and, in their work roles, they often have to choose between higher level ofsafetyand better performance (Abad et al., 2013; Caponecchia & Sheils, 2011). The differential value of productivity results associated with less use of personal protective equipment (PPE) and greater unsafe behaviors are associated with lower productivity levels (Arcury, Summers, Carrillo, Grzywacz, Quandt & Mills, 2014). For example, Arcury et al. (2015) report that Latino construction workers prioritize productivity over work safety mostly due to employment needs, comfort preferences and the created belief that their knowledge leads to better decision-making which minimizes potential work accidents. That is, from their experiential knowledge workers underestimate some risks (poor risk awareness) and overestimated own abilities. Therefore, it seems plausible to argue that workers may develop overconfidence due to the overestimation of their own actual capacities.
Nef is a 27/34 kDa, N-terminal myristoylated accessory protein involved in post integration infection. Nef is found in the viral particle and is one of the first proteins to be produced after invasion of the host cell. Although HIV-1 Nef was originally named "negative factor," it has been shown to have a positive role in viral replication and pathogenesis. Nef is a viral protein that interacts with host cell signal transduction proteins to promote long-term survival of infected T cells [148,149] and for destruction of non-infected T cells by inducing apoptosis. Nef also advances the endocytosis and degradation of cell surface proteins, including CD4 [12,13] and MHC proteins . This action possibly impairs cytotoxic T cell function, thereby helping the virus to evade the host immune response . The multifunctional protein thus helps the virus maintain high levels of viral load and to overcome host immune defenses, contributing to the progression of AIDS. Nef also alters the intracellular signaling pathways in lymphocytic cells, thereby inducing a wide range of effects. In particular, Nef activates both AP-1  and nuclear factor of activated T cell transcription factors , as well as the T cell receptor (TCR) chain signaling . Furthermore, Nef activates the calcium dependent signaling in T lymphocytes in a TCR-independent manner .
mutant. Virions were normalized by RT activity and used for transduction. Luciferase activity was measured 3 days post infection. (E) Immunoblot analysis of the expression and encapsidation of wt A3C and R122A compared to the respective Vpr-fusion-proteins. A3C constructs (+ or – Vpr) were detected by an anti(α)-HA antibody. Tubulin (Tub) served as loading control for cell lysates and p27 (capsid) for viral lysates. (F) RNA interacting with A3 proteins. A3C wt or mutant proteins and wt A3G were expressed in 293T and cell lysates were subjected to immuno precipitation. RNA bound to immunoprecipitated proteins was radioactively labeled with 32 P through RT-PCR and separated on a 12% PAA Gel and exposed on X-ray film. Background was set to signal of untransfected cells (mock). Equal amount of precipitated A3 protein was proven by immunoblot analysis of the elution fraction with an anti(α)-HA antibody. (G) RT-PCR on RNA interacting with A3 proteins. Isolated and A3 bound RNA (IP) was reverse transcribed and amplified using specific primers for 7 SL and 5.8 S RNA. Background signal was determined with RNA from untransfected cells (mock). Availability of the tested RNAs was confirmed for each sample through RT-PCR on RNA from cells prior to IP (cells).
Surgery and recording of baseline values
The abdominal cavity was approached via a midline laparotomy. The baseline tempera- ture and temperature after coagulation was measured at predetermined locations on the liver surface using a touchless standard infrared thermometer (VOLTCRAFT IR 260-8 S Infrared Thermometer; Conrad Electronics, Hirschau, Germany). At the same locations, the microcirculation was quantiﬁed using a combined laser Doppler ﬂowmeter and spectrophotometer system (O2C [oxygen-to-see] system; LEA Medizintechnik, Gießen, Germany). The temperature during coagulation was mea- sured with a thermographic camera (VarioCAM HD head 820; Jenoptik, ESW GmbH, Jena, Germany) on the ex-vivo liver surface for APC and HPC. This examin- ation was performed at diverse device output power settings ranging from 10 W to 100 W in diﬀerent areas. The infrared detector enables acquisition of infrared sequences with a spatial resolution of 1024 768 pixels with a temperature reso- lution better than 0.05 K. The camera was connected to a notebook via Gigabit Ethernet. Image sequences of 2-min length were recorded and analysed post-hoc using the IRBIS 3 plus software package (Infratec GmbH, Dresden, Germany).
The first heart xenotransplantation in humans ever performed was by James Hardy. He had been considering heart allotransplantation since he was inspired by the good outcome of some patients transplanted with chimpanzee kidneys. In 1964, Hardy decided to accomplish the first clinical cardiac transplantation and chose to acquire some chimpanzees as possible organ donors. He proposed a recipient who was in a semi-comatose state with extensive atheromatous vascular disease, for which he had taken amputations of both legs. Because the patient was deteriorating rapidly, Hardy made a decision to implant the chimpanzee heart. The heart turned out not to be strong enough to support the circulation and eventually failed within two hours. However, because the consent form for the operation did not mention that an animal heart might be used, Hardy's transplant surgery received opposition from the public and medical peers, so he did not perform any further attempt.
In lentiviralen Vektoren kann eine Kombination von Transgenkassetten zur RNAi und Überexpression erfolgen. Um viele Transgene platzsparend in einem lentiviralen Vektor zu kodieren, bietet sich beispielsweise die Verwendung von sogenannten 2A Elementen an (Donnelly et al 2001) wodurch ein Promoter die Transkription mehrerer Transgene regulieren kann. Die ca. 20aa langen 2A Elemente kommen natürlicherweise in Viren vor (F2A, Foot and Mouth disease virus; E2A, Equine rhinitis A virus; T2A, Thosea asigna virus) und stellen ein natürliches Spaltungsignal durch das ubiquitär exprimierte Enzym Furin (Furin cleavage Site) dar (de Felipe 2004). Viele Forschergruppen konnten unter Verwendung lentiviraler Vektoren die homogene Transgenexpression mit bis zu vier Transgenen verbunden über drei 2A Elemente nachweisen (Carey et al 2009, Donnelly et al 2001, de Felipe 2006, Fang et al 2005, Szymczak et al 2004). Chinnasamy et al zeigte jedoch , dass die Transgenexpression des letzten Transgenes auch geringer ausfallen kann, die Spaltungseffizienz Konstruktabhängig ist und fast 10% ungespaltene Proteine in den Zellen zu finden sind (Chinnasamy et al 2006). Auch können Deletionen bei großen Transgenen (de Felipe 2003) und fehlende Funktionalitäten der Transgene durch verbleibende Aminosäuren nach Spaltung der 2A Sequenz auftreten (Hasegawa et al 2007) was bei jedem Konstrukt und Transgen eine ausführliche Austestung in vitro nötig macht.
4.1 Uptake of dermatological relevant particles accumulations in the SC of healthy skin regardless of the shape of the AuNP. That is in accordance with the porcine skin investigations and with many publications which were not able to demonstrate skin penetration by NP (≥ 40 nm) unless in injured skin or in Franz diffusion cell experiments [83, 32, 6]. This is due to the massive barrier function of the SC which has been developing during the evolution and protection of human beings against all natural NP they were exposed to . The uptake experiments presented no penetration of the NP into the living skin layers in intact skin regardless of the material and size of the NP. But as silver wound dressings are used in cases of skin burns or other open lesions. The impaired skin barrier and particles might come into contact with the living tissue, where the uptake of AgNP into the cells is of interest. Therefore the penetration of spherical and prism-like AgNP into HaCaT cells and hMSC was investigated. STXM experiments presented a low uptake of the AgNP prisms into the HaCaT cells in comparison to the AgNP spheres. This could also be confirmed in TEM experiments by counting the cells containing AgNP (see appendix figure A.5). Nevertheless, the present work shows that the particles are taken-up by the cells. TEM demonstrated that the particles got into the cell after 24 h and were stored in endolysosoms. It was also shown, that hMSC do behave quite similar. The ratio of cells which take-up particles and those who do not was different for both cell types. The incubated hMSC showed a higher uptake compared to the HaCaT cells. However, atomic absorption spectroscopy (AAS) experiments to analyze the Ag content of cell extracts by Daniel Nordmeyer in his Phd thesis confirmed different amounts of silver in the two cell types comparable to the results of the cell counting . These results interestingly show that both cell types took-up the particles but hMSC presented a higher affinity towards the AgNP prisms. It can be concluded that the shape of the particles can play a role in cellular uptake.
PCMV belongs to the genus Roseolovirus in the subfamily Betaherpesvirinae in the Herpesviridae family . The enveloped viral particles are 150–200 nm in diameter and the only one completely sequenced PCMV contain a 128,367 bp linear double- stranded DNA genome with 79 open reading frames . Pigs are frequently infected with PCMV. The virus causes generally mild disease in animals younger than six weeks. Infected adult pigs may not present with symptoms of disease and the virus remains latent and therefore difficult to detect. The primary site ofvirus replication in pigs is in the nasal mucous glands or the lachrymal glands and the virus can be recovered from nasal and ocular discharge . Since the related human cytomegalovirus (HCMV) is an important cause of morbidity and mortality in human allotransplant recipients , it may be expected that PCMV has a similar effect. It is important to note that PCMV as Roseolovirus is closer related to the human herpes virus 6 (HHV- 6) and HHV-7, than to HCMV (HHV-5). In two studies it has been shown that PCMV induced a significantly reduced transplant survival when PCMV-infected kidneys were transplanted in two non-human primate species, baboons and cynomolgus monkeys [7,8]. In another case, in addition to the transmission of PCMV the baboon cytomegalovirus was activated also in the recipient , however, in all these cases it remained unclear, whether PCMV was able to infect cells of the recipient.
The finding that EPI-X4 also reduces the basal signaling activ- ity of CXCR4 and thus acts as an inverse agonist may have important implications because it suggests that CXCR4 activity is not solely dependent on CXCL12. CXCR4 is of significant inter- est as a drug target ( Peled et al., 2012; Cojoc et al., 2013 ), and EPI-X4 has interesting properties for clinical development because it mobilizes stem cells and prevents inflammatory cell infiltration in vivo. Moreover, it is not immunogenic or cytotoxic, whereas plerixafor, the only approved CXCR4 antagonist ( De- vine et al., 2008 ), affects mitochondria function ( Table S2 ). Furthermore, EPI-X4 is functionally distinct from plerixafor because it does not bind to CXCR7 and also acts as an inverse agonist of CXCR4. Many clinically used GPCR antagonists display inverse agonist activity ( Kenakin, 2004 ). Thus, EPI-X4 and its derivatives may be useful for the treatment of diseases associated with excessive activity of CXCR4, such as chronic in- flammatory disorders and WHIM syndrome, a congenital immu- nodeficiency associated with increased susceptibility to human papilloma virus infection ( Hernandez et al., 2003 ).
Similarly, the model can show why marketing activities by insurance companies should highlight the downside state. In particular, the use of salespeople who can vividly envi- sion the downside states might be an effective tool for insurance companies to increase demand. Initially, in a situation where she is not directly confronted with the potential downside state, and hence switching costs are low, a decision maker may choose an opti- mistic view by opting for frame (E). In turn, she would ignore the potential adverse states and abstain from an insurance purchase. However, an insurance salesman may succeed in making the downside state vivid which, in our model, could force her into frame (S), or alternatively, contact with the insurance agent may limit overoptimism by increasing the switching cost c S,E . Then, she may be unable to ignore the adverse states: using frame (E) might only go along with high psychological costs of avoiding thoughts on the potential downside risk. A decision maker may then be locked with frame (S) such that she over- weights the future loss states and therefore buys (even excessive) insurance. This distinct prediction on insurance purchases is empirically testable: people not in touch with in- surance sellers are hypothesized to be underinsured, whereas people tend to overinsure if they are directly in an insurance sales conversation. Hence, our model may provide a new rationale for the excessive use of sales agents, playing with the fears of decision makers, in the insurance industry.
Table 7: Allocation of owners in wealth groups, data and model
Note: The table gives the fraction of entrepreneurs that fall into a given category in the economy-wide wealth distribution. Data refers to net wealth from the SIPP.
Next, we turn to the model where we x owner hours at 40 per week. In this model, owners with high wealth still nd it attractive to operate highly productive rms, however, high wealth levels do no longer lead individuals to start low productive rms and work few hours. Instead, in the alternative model, individuals found low productive rms only when their productivity as a worker is even lower than their entrepreneurial productivity. Selection into entrepreneurship upon wealth allows our benchmark model to be consistent with median owner income being lower than the median worker income (0.76), yet, the median owner wealth being larger than the median worker wealth (1.4). Without selection upon wealth, in the alternative with xed owner hours, the wealth to income ratio of low productive owners, those who need little wealth to nance their operations, is almost the same as for workers. Therefore, though the model features an income ratio of the median owner to median worker similar to the baseline (0.74), it fails to generate a higher median owner wealth (0.8). As a consequence, the model also features too many owners in the bottom half of the wealth distribution (54%).