Background: Despite increasing popularity and mounting evidence for teacher collaboration as a lever for school improvement, reported changes in teaching associated with collaboration are often subtle and incremental, rarely involving substantial shifts in instructional practice called for by advocates of deeper learning and next-generation standards. One reason more expansive teaching changes remain elusive is that existing "horizons of observation" constrain possibilities teacher teams consider and solutions they develop while collaborating to improve teaching and learning. Purpose: This case study of two secondary school teacher teams explored the potential of collaborative partnerships with outside content experts (OCEs) for infusing new resources and perspectives that move beyond persistent images of classroom instruction. Setting: The study context was the Learning Studios model from the National Commission on Teaching and America's Future (NCTAF), in which interdisciplinary teacher teams partnered with local scientists and researchers to develop and implement yearlong project investigations with students. Research Design: The study used a qualitative case study design, including live observations and narrative transcription of team planning and design sessions, interviews, focus groups, and analysis of web-based interactions to develop rich narrative descriptions of the two partnership cases during the 2014-2015 academic year. We also compared and analyzed prominent patterns of OCE facilitative action across the two cases. Findings: Coding and analyses revealed several pivotal episodes of partnership interactions with clear evidence of OCE influence on teacher instructional plans. Cross-case analyses point to three OCE facilitative actions that preceded these effects--adapting expertise to local needs, following up between meetings, and judiciously applying pressure. Conclusions: The pivotal episodes we captured provide some initial evidence to support previous researchers' hypotheses that extended collaborative engagements can facilitate teacher learning in ways not readily achieved through traditional partnership models. The joint productive activity and depth of interaction observed in these cases opened up several opportunities to infuse knowledge and insights seldom documented in other teacher-expert studies involving loosely structured programs or short-lived externships. The two cases also provide initial evidence that outside experts can help to expand the horizons of possibilities teachers consider during instructional planning. Although these examples are instructive, we do not believe the reported effects are of sufficient magnitude or duration to be labeled "transformative shifts in practice." We also cannot say whether these initial changes in instructional plans translated into changes in classroom teaching or improvements in student learning. These remain important questions for future research.